Part Two

Governance Models in Systematic Design 12.2018


The first part of this dissertation explored how the relationship between visual languages and their significance plays a crucial role in establishing the foundation of a design system.

This second essay will explore the organizational structure and the governance models of communications systems and what are the implications that each one of them has in relationship to the semantic value the system generates (hereinafter also referenced as “data” or “information”).

Governance models

Governance models establish specific frameworks within which a communication system can operate.
These framework define the arbitration model, the presence of regulatory agencies, the directions information flows in, what are the acceptance standards, which parties own the data, how accessible information is and what are the points of failure inside of the system.

Stakeholder Identification

Before we dive into the various governance models, let's determine the semiotic stakeholders of a system. We can divide the actors of a communication system in two main categories: the authors and contributors (which constitute the regulatory agency, if present) are the ones in charge of creating and documenting semantic relationships between signs and significance.

The non-authors (or audience) is instead the group towards which the system and its conventions are propagated and is in charge of utilizing the semantic bonds within the guidelines imposed by the system itself.
Depending on the type of communication system we are describing, these two category sets can overlap by different degrees.

For ease of analysis, governance models will be first catalogued based on their arbitration model. Will will then look at how the other properties vary under different different governance models.

Closed Systems (centralized)

Closed communication systems (all systems which do not allow to be manipulated by actors outside of the author-contributors category), by definition have in place strong policies for unilateral information flows: data is shared in only one direction, when the authors of the system publish updates which are consumed by their audience.
Data ownership is limited to the authors and its often guarded by a sanctioning agency in order to preserve its integrity and grant for its correctness.

A diagram representing the arbitration model of a closed communication system: authors and non-authors are separated into different groups (A ∪ NA = 0). Its to be noted that the second group is often significantly larger than the other.

Closed communication systems tend to be very prescriptive and do not encourage the usage of their artefacts outside of the system’s guidelines. They are typically updated with large monolithic updates, over a very long period of time (decades).
This makes this type of systems very coherent, albeit slow and not particularly susceptible to change.

“One of the first modern-day road sign systems was devised by the Italian Touring Club in 1895. By 1900, a Congress of the International League of Touring Organizations in Paris was considering proposals for standardization of road signage. In 1903 the British government introduced four "national" signs based on shape, but the basic patterns of most traffic signs were set at the 1908 International Road Congress in Paris.In 1909, nine European governments agreed on the use of four pictorial symbols, indicating "bump", "curve", "intersection", and "grade-level railroad crossing”.
A brief history of traffic signs, according to Wikipedia

A typical example of a closed communication system are road and traffic signals. Most notably, the authors of this system are actors belonging to the Vienna Convention on Road Signs and Signals, which agree on a set of standards to be used in different countries.
The audience of this system (the citizens of the countries that adhere to that convention) have no say in the system - nor they are allowed to interpret the signs outside of the prescribed norms.

Open Systems (decentralized)

An open communication system is by definition a system in which its authors and audience coincide.
Since this system is created by the same actors who will be consuming it, it has a strong bilateral feedback loop, which encourages continuous modifications to its semantic value and a maintains the system in a constant state of “never finished”.

In an open system, authors and non-authors overlap, forming a union between their sets (A ∪ NA > 0).

The largest open communication system in existence is the linguistic one: a construct in which all actors are actively (even though sometimes unknowingly) participating in creating and propagating new semantic relationships between words and their meaning, as well as consuming the constructs that are created and propagated by other actors.

The acceptance model in this case is purely based on velocity and critical mass. Constructs are distributed at very high speed and once an expression becomes part of the oral linguistic system, it becomes a candidate for inclusion in the Merriam-Webster dictionary (a example of a selectively open system).
This shows how semantic value does not live and die within a certain governing body, but is able to move between different models, over time.
Despite the volume, intricacy and number of actors involved, language evolves at greater speed than most systems due to its non-existent barrier of entry.

  • Use of "adulting"
  • Use of "FOMO"
A scatter graph showing the raise in popularity of the words "adulting" and "FOMO" on Reddit between 2015 and 2018 [data from How The Internet Talks].

Selectively Open System (federated)

Selectively open systems maintain the separation between authors and audience as discreet entities, but they allows for co-authoring upon fulfillment of certain criteria.
In this model, only a number of selected contributors (usually chosen based on their reputation or expertise in the field) are part of the governing agency allowed to contribute the system, while all the other actors only consume the system.

A diagrammatic representation of a federated model. The sanctioning group is represented by the intersection between authors and non-authors (A ∩ NA > 0).

In addition to the aforementioned Merriam-Webster, another example of federated communication system is the W3C consortium.
The W3C is a members-only organization that discusses and consolidates standards that affect billions of users, but that is comprised of only 476 members, plus a regulating organ which formalizes the decisions made by the members, based on consensus.
Federated models can have different acceptance standards, but they usually require peer review and a variable degree of agreement.

Comparing attributes

We started this essay by listing some of the properties which differ across governance models.
Its now time to look at each one of them comparatively, so that we can better understand the ramifications of working within each of the models.

Governance Models
Arbitration Model
Regulatory agencies
Information flow
Acceptance standards
Data ownership
Data accessibility
Points of failure
Closed Systems
Strictly regulated
Open Systems
Critical mass
Selectively Open Systems
Selectively bilateral
Matching quality & criteria
Centralized & redundant
The table above documents the values for the individual attributes across the three governance models examined earlier.

Often governance is thought only from a gatekeeping standpoint, but there's much more that is involved when establishing the arbitration structure of a system.
Information ownership, for example, can be an important issue when dealing with closed system, where only the arbitrating body has write access to the data.
Conversely, redundancy decreases in a closed model, making the systems which adopt this type of architecture more susceptible to failure than an open system, which instead can rely on a completely distributed framework.

Wrapping up

Understanding the implications behind a certain governance model is crucial for the success of the system and its adoption.
Information naturally aspires to be distributed freely so its important to be cognisant of the practices we put in place to control its circulation and make sure they don't have any unintended consequences.

A heavily bureaucratic system will move slower than a completely distributed one, and risks to get out of sync with the ever-changing needs of its audience. On the verge of becoming obsolete non-authors quickly abandon the failing system and leapfrog quickly to adopt a new one.


Subscribe daily or weekly to the most relevant industry links across design, product, and development. Published via