Passau, Germany — on the Beautiful, Blue Danube

At some level, it’s instructive to go back and review the longer definition of a Heuristic, and then consider what the implications are behind NOT going with a particular algorithmic approach.  Academics typically get conniption fits with heuristic thinking — usually along the lines of NOT RIGOROUS ENOUGH.  Too touchy-feely — which is really a term for a poor understanding of more evolved empathy!

Some of this might be a manifestation of v-Meme conflict that we’ve covered earlier.  The non-empathetic don’t particularly like, nor understand more empathetic approaches.  And if the stretch is long enough, we’ve already covered the fact that one can run into major hostility.  But do they have a point?  And if they do, what can we do to remediate their concerns of rigor?

The definition from Wikipedia is as follows —  A heuristic technique (/hjʉˈrɪstɨk/; Ancient Greek: εὑρίσκω, “find” or “discover”), often called simply a heuristic, is any approach to problem solving, learning, or discovery that employs a practical methodology not guaranteed to be optimal or perfect, but sufficient for the immediate goals. Where finding an optimal solution is impossible or impractical, heuristic methods can be used to speed up the process of finding a satisfactory solution. Heuristics can be mental shortcuts that ease the cognitive load of making a decision. Examples of this method include using a rule of thumb, an educated guess, an intuitive judgment, stereotyping, profiling, or common sense.

Often, heuristic methods involve ‘messing around’ with some problem (experiential learning) and embodying the ‘fail early, fail cheaply, fail often’ mentality we introduced in the last post.  That’s not going to make the Power and Control, or the Rule-Following crowd very happy.  With such processes, it looks like there is some indeterminate end, or even worse, an end determined in some fashion by the participants in the process.

And if you scroll back and remember how timescales are calibrated by empathetic development, now you’re bringing on the Major Crazy to the Authoritarians (time scales are decided by the authority above them) or the Legalists (time scales follow rules set from the outside — remember the creation of time zones/railroads story.)  The people involved in the process aren’t supposed to have the agency to set the limits on time for a project.  That’s above their pay grade.

What is needed by the Performance v-Meme crowd (get to the goal) that will mollify those lower down on the v-Meme Spiral is some kind of heuristic design process that resembles an algorithm, or at least a plan.  And then dependent on the technical requirements of the process, this plan appropriately scaffolds analysis effort into the design, in the quest for validity, with enough reliability to move forward with confidence.  These naturally fall out of the different Spiral v-Meme levels for a given project.

What might those be?  Just like a Matryoshka doll, higher/larger levels must be filled with the lower levels.

Matryoshka dolls, from Wikimedia

1. Legalistic/Absolutistic — enough analysis steps and algorithms (think computer-based finite element analysis, cursory flow analysis, statics, thermodynamics, etc.) that the potential designs do not violate the laws of physics.
2.  Authoritarian — enough facts, figures, and understanding prior art that a design is appropriately referenced to what has come before.
3. Tribal/Magical — referencing prior knowledge and stories that exist in the organization on past projects that have created the iconography behind specific choices.  This last one, in many ways, is the toughest, because of past failures with particular technologies.  A deeper dive will be necessary to understand the story well enough that changed conditions and technology development arcs will create an argument that will allow larger change.  A great example might be the adoption of lithium-ion batteries in the Boeing 787, and the resultant battery fire.  Clearly, lighter batteries are going to be used in aerospace applications.  The challenge is to assure reliability to commercial aviation standards.  Yet the experience was so traumatic (potentially nothing is worse than a fire inside an aircraft) that the story has been encoded into Boeing’s collective memory as the threshold statement for a major crisis!
4. Survival — the circumstances inside the company RIGHT NOW are amenable to a given design process.  This likely means sufficient budgeting, no other ongoing crises demanding the attention of group members, and so on.

Modern Computer-Aided Design and Analysis tools have made addressing #2 even easier than before.  Reliability of such tools has increased that a minimum of training is necessary for cursory results, and the acceptance that this is actually the case has spread even in the hierarchy of engineering schools.  It used to be that, for example, stress analysis using computers was taught as a graduate class, and there was much consternation about error control, meshing, and such.  Now, freshmen engineering students in our drafting class are taught the basics of how to find out the stress on a part given a loading condition, and the analysis tool itself is used to develop heuristics inside students’ heads for material behavior!

So what might a more structured heuristic design process look like?  That will be the subject of the next post.  But before we get carried away, I think it is very important to note: we’ll only discuss one potential heuristic.  Like all products of thought, it evolves out of the social structure that created it — this is absolutely inescapable, until we obtain some level of self-awareness on why we think the way we do.  In future blog posts, I’ll discuss other design heuristics and methodologies that are reflective of exactly those same types of mental dynamics, but at a higher v-Meme level.  And then, finally, we’ll wrap up with some speculation on how understanding higher-order connectivity can help us design the multiply connected, synergistic systems of the future.

Takeaway:  You can’t just wing it when it comes to the design of complex, technological systems.  You have to provide appropriate scaffolding that recognizes past Tribal Knowledge, known facts, the Laws of Physics, and whether there are enough donuts on the counter.  Because the Second Law of Thermodynamics is the law.  And if people are hungry, they won’t pay attention!

Credit where credit is due:  Lou Agosta, of the Chicago Empathy Project, pointed out to me that using the matryoshka analogy for empathy didn’t start with me.  Franz De Waal, in his seminal text, The Age of Empathy, he has equivalent mapping in the bottom three levels of his empathy model, and rightly deserves credit for being first to talk about the nested level of the bottom three levels of my Empathy Pyramid.