Some Reflections on Principles
2023 Jan 27Some Reflections on Principles
This is an encore of an essay originally published in 2021 on Medium
Some definitions and reflections on their application in sociotechnological systems.
This essay (or attempt at one) is a collection of reflections and collages gathered over approximately 7 years of my professional life, spent in organizations with countless fundamental principles and others with absolutely no operational principles whatsoever.
My goal was never to “write this essay” in any definitive sense, but rather to close — even if only partially — this topic in my head through observation, reflection, empiricism, and speculation about how organizations with no principles are, by default, chaotic and dysfunctional.
If you’d like to stop reading here, I’ll leave you with a single partial conclusion from what I’ve thought about over this time:
Organizations structured around solid and explicit principles have a higher degree of cohesion among people, fewer conflicts over fundamental issues, greater consistency in decision-making, and correct themselves at a faster rate.
I’ll try to pull the topic toward technology as much as possible, but first — what exactly are principles?
What are principles?
There are many definitions out there regarding what principles are, some leaning toward a more etymological sense and others carrying a heavier semantic weight.
Some definitions of principles:
-
According to Wiktionary, the word “principle” comes from the Latin prīncipium, meaning “beginning” or “foundation.” As a noun, it equates to a “fundamental assumption/essence,” “faculty,” or “rule or law of nature” that establishes the basic idea of how things are applied.
At the risk of abusing semantic oversimplification, I see principles in technology as a set of assumptions, norms, and/or core beliefs that serve as the rational foundation to guide any and all actions within a sociotechnological system.
These principles serve as a kind of compass or reference point for grounding behaviors and choices, by establishing a shared awareness among the people within that sociotechnological system.
The point I want to make here is that principles will, ultimately, arbitrate the practices and choices of a given collective of people.
And those choices can range widely — from answering “What technology stack should we use for project X?” to “What behaviors do we intoleranly reject within our IT team?”
With this, all practices gain a slightly more objective character, because there is a north star, a direction, a foundation from which something is being done. And here, arbitration becomes much simpler and more direct for everyone.
But before moving to the technology examples, I want to illustrate the power of principles in a technical engineering domain.
Heavier than air…
To illustrate the power of principles in engineering and technology, I’ll use the activity of aircraft flight.
If we think broadly about the principles governing aircraft flight, there are 4 fundamental principles that will guide every activity in aircraft design, engineering, and construction:
-
Aerodynamic lift — the vertical perpendicular force that counteracts the weight of the flying object and keeps it airborne;
-
Drag — the aerodynamic force created in opposition to the object in flight;
-
Thrust — represented by positive propulsion; and
-
Weight — established by the gravitational force that “pulls down” all objects within Earth’s gravitational field.
In a hypothetical scenario where we need to build an airplane, no matter how much we might debate aspects such as project design, operational capacity, aesthetic elements, cockpit philosophy, and so on — by virtue of these fundamental principles, the sociotechnological system building that aircraft will have to comply with these principles in order to produce an airworthy aircraft.
Period.
Ultimately, an aircraft — to be minimally airworthy — must adhere to the principles; otherwise:
-
An aircraft unable to generate aerodynamic lift cannot even leave the ground;
-
A wing design with laminar flow problems can increase drag excessively, destabilizing the aircraft and threatening flight safety;
-
An engine without the capacity to generate sufficient thrust won’t give the aircraft adequate positive airspeed; and finally,
-
An aircraft that is far too heavy for its operational capacity won’t take off, and one that is too light becomes more susceptible to storms, turbulence, gusts, and wind shear.
Without fear of committing a logical leap with such a blunt example, we can see that (i) there is an interrelationship between principles — none is a sterile, disconnected unit, but rather part of a synergistic set of forces — and (ii) that violating one or more principles renders the system itself, by definition, either (a) dysfunctional or (b) unviable.
Principles: To have or not to have?
I’ve had experiences in technology teams at both extremes — some with very well-established operational principles, and others with a complete lack of any minimum operational principles.
Comparing organizations with explicit operational principles to those without, at least from my perspective, some clear differences emerged:
-
While organizations with operational principles managed to get most people to make objective judgments on fundamental matters — such as choosing a technology stack — their counterparts without principles always had elements of subjectivity and arbitrariness in practically everything, often with people choosing whatever technology would look best on their own resume, disregarding the trail of destruction and costs it would leave behind;
-
Some companies knew how to leverage the flexibility of their principles, since even when established they still allow a spectrum of interpretation for their application. Meanwhile, companies without principles had as their norm infinite ad hoc rules (usually arbitrary) that rigidified every practice and invariably converged toward micromanagement (e.g., companies tolerating toxic behavior in code reviews or harassment); and
-
In companies with principles, people most often had predictability in relation to each other’s behavior, and doing the right thing was always expected; while in companies without principles, insecurity was the prevailing sentiment, since the possibility of some arbitrary action from any member of that sociotechnological system was always on the table.
Like our aircraft example, an organization without principles becomes entirely dysfunctional — because without the central ideas of how something should work, anyone can decide anything and, in the end, everyone will be right.
Idealized and defensive principles
I believe no organization establishes its principles simply by chance or by referencing some external source.
A characteristic of organizations that grounded their operations in very well-defined principles is that these organizations formed their principles as a patchwork quilt containing values, experiences, present knowledge, perceptions, heuristics, and very unique visions of how something does or should work.
But no matter how unique all those principles were to each organization, they all fell into two categories: the first being idealized principles and the second defensive principles.
Idealized principles are related to a vision of the future and how adherence to them would lead toward an ideal path for reaching a particular place and/or something specific. That “something specific” can range from operational excellence to the establishment of a new quality standard.
An example of idealized principles are Apple’s design principles.
These principles hold that every human-machine interface experience provided by their devices must follow principles such as simplicity, minimalism, focus, precision in detail, aesthetic integrity, intuitive user control, and interface consistency.
We can notice that these principles sound extremely positive, as they evoke the potential pinnacle where, by following them, their interfaces will deliver the best user experience on their devices.
In turn, defensive principles are those conceived to safeguard the organization if something goes wrong — or in some extreme cases, defined after something goes catastrophically wrong.
Here we find numerous principles such as reliability engineering for resilience guarantees, technology choice directives (and here too), safety principles for critical industries like aviation where things going wrong means people simply die, or even privacy principles in industries where customer trust is the business model.
FINAL THOUGHTS
At another opportunity I’ll speak a bit more about defensive principles, particularly in situations involving the prevention of catastrophic events and how this can be both good and bad.
One case in which a defensive principle with a correct rationale was decisive in avoiding a failure: I worked at a company where Windows was banned for SysAdmins and developers with server access, on the grounds that the company’s principle was not to use closed-source code. This principle saved the company from the 2017 Ransomware attack — while 90% of our partners were offline, the company didn’t experience a single second of interruption.
The opposite situation also occurred — a defensive principle that went very wrong involved a technology team that moved all applications to the cloud after years of reliability problems with on-premise data centers. The result was that due to a poorly written contract and political instability, the company paid a cloud bill in US dollars at BRL 3.98 + taxes, when the exchange rate at the time the contract was signed was BRL 2.27 — not to mention the forced migration of everything from the cloud provider back to the old data center.