Defining Stigmergy Part 1

Something I keep coming back to is 'how far can/should the term stigmergy be extended?' The concept in general seems applicable to so much of human endeavour that I am concerned that without focusing, the term risks being watered down to the point of ineffectuality.

The original definition given by French biologist Pierre-Paul Grassé in 1959 to refer to termite behaviour was: "Stimulation of workers by the performance they have achieved." From the broadest interpretation of this definition, culture itself seems to qualify as stigmergy - human workers modify our surroundings and in turn are stimulated to do more and better. However if one is to apply the term in a focused and useful manner, such a broad definition is no doubt problematic in that it leaves little scope for such distinctions as stigmergy on the internet vs. in culture etc. Never mind the fact that all organisms stimulate their brethren through the performance they have achieved at least in evolutionary terms.

So it should have come as no surprise after doing some hunting around that many researchers have addressed the issue of defining stigmergy. I discovered a nice little Power Point poster by Dylan Shell and Maja Matarić: On the use of the term ‘Stigmergy’ (.ppt). (Here's an HTML version, however it lacks all the references.) This poster shows that not only is the definition up for contest, with confusion over whether communication takes place indirectly or at all, but so is the term's etymology. Below are some variations on the terms origins (see the poster for references):

• "‘stigma’ (goad) and ‘ergon’ (work)…” to “…stigmergy (stigma: Wound from a pointed object; ergon: Work, product of labour = stimulating product of labour).”
  
• “from the Greek words stigma ‘sign’ and ergon ‘action,’ and captures the notion that an agent’s actions leave signs in the environment, signs that it and other agents sense and that determine their subsequent actions.”
  
• “[F]rom the Greek stigma: sting, and ergon: work.”
  

Perhaps some of the more interesting research done with regards to stigmergic collaboration that is highlighted in the poster is the various categorisations thus far developed in research:

Holland and Melhuish [1999] define two subtypes:
Active: actions affect sensory inputs of others
Passive: actions affect the outcome of later actions

Bonabeau, et al. [1999], distinguish:
Qualitative: sensory inputs differ by type
Quantitative: sensory inputs differ by degree

Wilson [1975], draw the distinction between
Sematectonic: change in state result of task-related action
Sign-based: (or marker-based) result of something that makes no contribution to the task

As the authors of the poster point out, Grassé's original understanding of stigmergy then becomes further defined as: Active, Qualitative Sematectonic stigmergy - that's a mouthful. More on this in Part 2.