Social entity thinking wrt EA

Copyright 2020 Graham Berrisford. Now a chapter in “the book” at Last updated 27/10/2021 17:45


This chapter addresses some points relevant to the success of business entities, notably principles to be followed, aims to be achieved, and structures for organizing interconnected human actors. It introduces only a fraction of what might be said about organization design and business management, focusing on ideas most relevant to system theory and to enterprise architects.


Reading online? If your screen is wide, shrink its width for readability.




Preface (recap) 1

Social entity thinking as a branch of system theory. 1

Defining general principles for actors. 1

Defining specific aims for actors. 1

Defining how actors are “organized”. 1

People factors. 1

Conclusions and remarks. 1

Appendix: Some famous sets of principles. 1


Preface (recap)

In defining aims, the normal practice is to zoom out, encapsulate the system of interest, and consider the effects it should produce. The overarching concern is the desired outcomes of system activity.


In defining the activity system needed to meet aims, the focus is on defining processes that produced desired effects, the rules to be followed and roles actors play in activities.


In defining the social entity in which actors are employed, the focus is on how actors are directed, motivated and organised to perform required activities; including general principles for actors to follow, specific aims for their activities, and the structure under which actors managed.


Whether your focus is on the activity system or the social entity depends on where you are coming from and what you aim to achieve. You may flip from one viewpoint to the other, but they come with different ideas about what it means for a system to be defined, to exist, and to change or evolve.


It has been said that EA regards an enterprise as a "system of systems". More accurately, EA sees a business as a social entity that employs a messy patchwork of activity systems, which may interact, overlap, and even be in competition. EA strives to extend and improve these activity systems, and optimize how they are coordinated to the benefit of the whole.


No business is a wholly distributed collective of independent agents. To some extent, the actors must be motivated and managed to act as business directors wish them to act in the performance of more or less regular business operations. So, EA also requires a measure of social entity thinking. At the very least, logical functions and roles must be mapped to physical organization units and actors. The advice in short is:


Is your interest where actors

Then use

determine their own actions?

Social entity thinking


do both above and below?

Social entity thinking

Activity systems thinking

play roles in regular processes?


Activity systems thinking


This chapter focuses on social entity thinking, and its relevance to EA.

Social entity thinking as a branch of system theory

General system theory didn’t start from sociology, or analysis of human behavior. However, it stimulated people to look afresh at social systems and business systems. In 1956, Kenneth Boulding wrote on what was becoming known as management science. In his article, he presented general system theory as the skeleton of science. He asked what are the elements of a social system? Are they actors, or the roles actors play in activities?


This question has hung over social system discussions since the nineteenth century.

"The first decision is what to treat as the basic elements of the social system. The sociological tradition suggests two alternatives: either persons or actions." Seidl 2001


In short, the centre of your thinking, you may put either the actors or the activities.


People focus

Social entity thinking

Action focus

Activity systems thinking


A network of actors, who communicate and act as they choose,

A network of regular activities, performed by actors.


A card school.

A poker game.

How to scope it?

The boundary is drawn around an identifiable group of actors (a species, a tribe, a card school, the employees of a business).

Almost regardless of what activities they perform.

The boundary is drawn around those activities needed to maintain some variables. and/or produce some (emergent) effects or results.

Almost regardless of what actors perform the activities.

How do actors determine their actions?

Actors may be more or less free to invent how they act, bearing in mind given aims, shared aims and individual aims.

Actors select from the range of regular activities allowed by the system.


Importantly, the relationship between activity systems and social entities is many to many. One activity system can be realized by several social entities (the game of poker can be realized by many card schools). Conversely, one social entity can realize several activity systems (the card school can play whist, or share a pizza).


In the table below, activity systems thinking focuses on the higher rows, whereas social entity thinking focuses more on the bottom row.


How an abstract activity system is realized

Orchestral music

Game of poker

Abstract system: a description of roles for actors, rules for processes and variable types

A musical score

The rules of the game

Physical system: a performance of defined activities, which gives values to variables

A performance of the score

A game of poker

Physical entity: one or more physical actors able to perform the activities

An orchestra

A card school

Causality and choice

Management science addresses human institutions that – in more or less bureaucratic ways - organize some human actors to do something. Typically, these social entities employ many discrete activity systems, in each of which the range of possible actions is defined. You rely on countless such activity systems; for example, you wouldn't want to:


·       stand trial in a court that didn’t follow court procedures

·       board a train or airplane operated by people who didn’t follow the rules.

·       loan money to a company that didn’t repay its loans as promised

·       play poker with people who ignore the laws of the game.


When observing actors’ behavior, we can classify their responses to stimuli into four kinds.



In theory, when an event happens, we can predict


exactly which action an actor will perform in response.


how likely an actor will perform activity type A or activity type B.


the actor will choose from the range of activity types in our model.


nothing – because actors can invent activities outside any model made.


In so far as a social entity realizes a known activity system, its actors always act in the first three ways. We may not be able to predict which action they choose to perform, but we can say they will choose one of the actions available to them in the activity system. (A designer has to make allowances for “exception paths” actors may choose to follow.)


At one extreme (think of a rule-bound social entity like an ant colony) only the first three kinds of causality are possible. Expanding the range of possible actions gives actors a higher degree of freedom, and increases the system’s complexity, but it remains describable as an activity system.


At the other extreme, only the fourth kind of causality applies. Where every action is self-determined, where there is no observable regularity, repetition, or pattern, then there is no recognizable activity system in the social entity.


(When actors act in the fourth way, they act outside any activity system we know of. Even if they are in fact in acting in an activity system we don’t know about, we must treat them as having free will and the ability to do what they choose.)


Surely all real-world businesses sit between those two extremes? It is impossible to model every kind of message an actor can send or receive, and every action they can usefully perform. Only some activities can usefully be regularized. The extent to which activities are best “organized” varies from business to business. Directors expect employees to have insight and determine actions that are necessary or helpful to meet the aims of the business.


In any case, to a greater or lesser extent, actors must be motivated and managed to act as the business directors wish them to act. Social entity thinking goes some way to direct actors without actually prescribing or triggering specific activities. The rest of this chapter focuses on defining:


·       general principles for actors to follow

·       specific aims for activity, and

·       the structure in which actors managed.

Defining general principles for actors

Where actors are independent agents, more or less free to determine their own actions, it is common to provide some principles or guidelines, which steer the actors towards making one choice rather than another.

Design pattern pairs

In software architecture there are several design pattern pairs, four of which may be contrasted as distributed versus centralized.


Distributed pattern

Centralized pattern

Network connectivity

Hierarchical connectivity

Point to point communication

Hub and spoke communication

Peer to peer cooperation

Client sever cooperation

Choreography/chain control*

Orchestration/Fork control


Choreography/chain is where an end-to-end process is completed by hand over from one actor to the next, with no overarching controller. It does not imply the actors are free to act as they choose.


In software system design, the designer’s role is to understand the trade-offs, mix and match the patterns to the situation, and prioritise what is practical.

Social entity thinking values and principles

Some see the lesson of systems thinking as being "everything is connected and we've all got to work together to [insert goal here]", and promote principles of the kind that promote the concepts to the left in this table.


Distributing principles

Centralizing principles

Network connectivity

Hierarchical connectivity

Point to point communication

Hub and spoke communication

Peer to peer cooperation

Client sever cooperation

Self-organization of actions*

Direction and coordination of actions


The term “self-organization” here can convey various meanings. To begin with, let us assume it means actors are responsible for their own actions.


Imagine the directors of several charities meet to discuss their concern that no one of them can meet the multiple needs of their clients. Each organization was founded to tackle specific needs and specializes in those. How to be customer-centric to the point of addressing every need that each citizen has? Out of their discussions, three options emerge.


·       Option 1: A central government department assigns every citizen to a coordinator, whose job is to engage all organizations equipped to help that citizen.

·       Option 2: Distribute that responsibility to employees in each charity’s organization, who must be trained and directed to identify where a citizen has needs addressed by other organizations, and engage them.

·       Option 3: Distribute that responsibility to the citizens, each being an agent empowered to contact the organizations from which they seek help.


A naïve application of the principles in the table above might lead the directors to favor option 3. In practice, they are more likely to favour the option they have most control over – option 2. More generally, we should look for the optimal solution to a system design problem, rather toward some extreme ideal.


In EA? EA principles mostly relate to the architecture of activity systems, rather than social entities. There is however considerable pressure for architects to adopt “agile” principles.

Values and principles for agile design and build projects

For decades, people have been promoting agile software development project principles, and some have looked to apply them more widely. In practice, managing a project to design and build something new is one thing, and managing regular business operations is another, but let us a look the former.


Agile airplane design principles

In the second world war, Lockheed Martin recognized the needed to design new airplanes as fast as possible. The manager of what was called “Skunk works”, called Kelly, documented 14 rules for how to do this. Among his rules (see appendix 1) you can find four that correspond to agile software development principles today.


Kelly’s rules for Skunk works – 1943

Use a small number of good people.

The contractor must test his final product in flight [and] in the initial stages.

A minimum number of reports required, but important work must be recorded thoroughly.

There must be mutual trust between the [customer] and the contractor [and] very close cooperation and liaison on a day-to-day basis

A very simple... release system with great flexibility for making changes must be provided.


See appendix 1 for Kelly’s 14 rules.


Agile software development values and principles

The agile manifesto for Agile software development projects promotes some social entity thinking principles. “We are uncovering better ways of developing software by doing it and helping others do it… while there is value in the items on the right, we value the items on the left more.”


Agilists value


Individuals and interactions

Processes and tools

Working software

Comprehensive documentation

Customer collaboration

Contract negotiation

Responding to change

Following a plan


See appendix 2 for the 12 principles in agile manifesto, which include collaboration, self-organization and continual learning.


The principles and values of “Scrum” (as in the Scrum guide of 2020) have more to say about the sociology of a software development team.


·       Scrum helps a group (who collectively have and share the skills and expertise needed) to solve complex design problems.

·       A scrum team is typically 10 or fewer people, on the basis that smaller teams communicate better and are more productive.

·       Scrum is based on the ideas of empiricism (knowledge comes from experience and observation), lean thinking (reduce waste and focus on the essentials), and an iterative, incremental approach (to optimize predictability and to control risk).

·       Scrum’s pillars are transparency (the process and work must be visible), frequent inspection of artifacts and progress toward agreed goals) and adaptation of process and artifacts, the moment things change.

·       Scrum’s values are commitment, focus, openness, respect, and courage.


When are agile principles most naturally applied?

A considerable amount of social system thinking discussion is influenced by agile software development principles. But in system design, all is trade-offs. It is important to realize that agile development methods grew of experience of what works well in certain kinds of project, not all projects. This table gives you a way to score a software development project for the suitability of agile methods.


What kind of project?


0, 1, 2, 3

Mandatory requirements-driven

Users available for Joint App Development

0, 1, 2, 3

Users not available

Developers empowered

0, 1, 2, 3

Developers not empowered

What kind of system and work?

Divisible into usable releases (soft target)

0, 1, 2, 3

Indivisible (hard target)

Simple UX and/or UI technology

0, 1, 2, 3

Complex UX and/or UI technologies

Output/enquiry/report dominated

0, 1, 2, 3

Input/update dominated

Simple (CRUD) data processing rules

0, 1, 2, 3

Complex data processing rules


0, 1, 2, 3



0, 1, 2, 3

Highly integrated w other systems

Add up the scores for your agile potential quotient


A project that scores less than 13 is likely to find agile development principles are readily applicable. The higher a project scores above 12, the harder it is to apply agile development principles. That doesn’t make the principles wrong, and you may still apply them as best you can. It probably does mean the project will be relatively challenging. You may have to take an approach that is more “waterfall” in nature, and requires more by way of up-front design. And relatively speaking, it may cost more and take longer.

Values and principles for regular business operations

Most regular business operations are very different from software engineering projects. Consider working in a call centre, digging holes in roads, performing operations in a hospital, manufacturing bicycles, or working in a bank.


In the 1970s, it was widely perceived that business managers in the USA had a lot to learn from managers in Japan’s motor car industry, and a “transformation” was needed. In “Out of the crisis: quality, productivity and competitive position” (1982 & 1986, Cambridge University Press, Cambridge) Deming promoted 14 points which (it is said) had been applied by business managers in Japan from the 1950s. See appendix 3 for Deming’s principles.


In EA?

This chapter illustrates some principles for actors in social/business entities who can, to some extent, determine their own actions, and perhaps even aims. By contrast, most enterprise architecture principles relate to options in the design and build of regular human activity systems and IT systems.


What is the difference between a principle and an aim? A principle is a general guideline. An aim is a target to be achieved that should be Specific, Measurable, Actionable, Realistic and Timebound (SMART).

Defining specific aims for actors

“A set of actors and resources that is coherently organized and interconnected so as to perform a characteristic set of activities, to achieve some aims". Meadows


A business may be defined as a social entity that realizes one or more definable activity systems, to meet some shared aim(s). Systems thinkers often use the term purpose. Trouble is, purposes are variously discussed as reasons to do things, uses we find for things, and outcomes we observe.

Actual outcomes (outcomes)

When talking about a natural system, we often speak of its actual outcomes, results or effects as its aims or purposes. Stafford Beer, Meadows and other system thinkers have gone so far as to say: "the purpose of a system is what it does" (POSIWID). Does this make sense?


Teleology is the explanation of phenomena in terms of the purposes they serve, rather than of the causes by which they arise. Macaque monkeys (in Thailand) have learnt to use stones to open oysters and nuts. Is the existence of the stones explained by the uses that monkeys make of them?


“Biologists do speak of the purposes of the heart, or the liver, or a kneecap. But that is just a way of speaking. The teleological metaphor was just a metaphor: underneath it lay quite simple mechanical explanations. Today’s scientists are pretty certain that the problem of teleology at the individual organism level has been licked. Darwin really was right."


In EA?

Everything we say, do or create can have unintended consequences. To speak of those outcomes as purposes seems perverse. In EA, the primary interest is in the desired outcomes of a new or changed system.

Desired outcomes (aims, goals, objectives)

Some animals (not only humans) display consciousness of their environment. Consciousness gives them the ability to think about the past and envisage the future, to have intentions, to set aims and act to achieve them.


In talking about a designed system, we usually refer to its desired outcomes, results or effects as its aims or purposes. These desired outcomes, motivations or goals, are ascribed to a system by external or internal actors.


Much as Meadows put it, a powerful way to influence the behavior of actors in a social entity is through changing its desired outcomes. Changing these aims may require people to change the activity systems they play roles in. (Actual outcomes can and should be compared with those, after a change has been made – but often aren’t.)


In EA?

There can be several levels of scopes of aims, goals or objectives. There can be aims for a business as a whole, aims for an organization unit, aims for a particular activity system, and aims for a project that is changing one or more activity systems.


Who declares the aims of an organization, how they do it, and how far individual employees are involved, are all influenced by the nature of the business.


Bottom-up? Systems thinker Peter Senge recommends building a shared vision from personal visions through interaction, give and take. In the course of that process, some may be obliged to suppress a personal aim or vision (that others in the same social entity don't share or don't accept). And note that people may belong to different social entities with conflicting aims and visions.


Top-down? Many business goals define something akin to a "balanced score card" via which goals are cascaded downwards through the organization structure. This approach has not been universally successful. Sometimes, aims declared or documented by business directors don’t entirely reflect their true aims. Sometimes, goal definition becomes an exercise in how to ensure bonuses can be awarded, or not awarded. Sometimes, it takes so long to cascade the goals they are out of date by the time they reach the workers.


Formal top-down goal definition is not presumed in EA. But it is expected that business directors identify and define the overarching business mission, vision, drivers and goals. And the goals of each line of business, team and project can and should be explicitly related to whatever principles and goals the business directors have declared.


Quantifying aims

In business, as in system dynamics, what at first may be expressed as qualitative goals are better cast as measurable quantitative variables. For example, “Improved Health" and “A Better World” are not meaningful goals. “Improved” and “Better” require there to be an ordering, and so, require quantification.


“When you can measure what you are speaking about, and express it in numbers, you know something about it, when you cannot express it in numbers, your knowledge is of a meager and unsatisfactory kind; it may be the beginning of knowledge, but you have scarcely, in your thoughts advanced to the stage of science.” Lord Kelvin


Of course, choosing the right variables to measure is challenging. Measuring variables is challenging. And responding to naively set, or inaccurate measures can be dangerously misleading It is important to focus on the right variables, rather than ones you can most easily measure. A concern about this was expressed by Hayek in his famous Pretence of Knowledge” speech.

“Fooled into believing that the [readily] measurable variables are the most critical, economists propose “solutions” that actually worsen the problem.” Hayek


Moreover, once people know how they are being measured, they may adapt their behavior so as to make the measure useless.


Qualitative aims

In system theory, the requirements for a system can be expressed in terms of outputs or state changes. In business, these appear in the form of product or service specifications.


The term "service" reflects the economic shift from the physical to the abstract. From the industrial age to the information age. From material processing to information processing.


A service is a time-bound activity that provides a result of value to a consumer. The result can be material (food, clothing), information (money, music) or other (social care, shiny shoes). In scientific terms, the result is an output of or state change in the system of interest. In business terms, a service is Specific, Measurable, Actionable, Reality and Time-bound.


Services can be coarse-grained or fine-grained. Macro service examples include, build a house, complete a project, place all this year's applicants in universities.


In TOGAF 9.1: "A service is a logical representation of a repeatable business activity that has a specified outcome". In ArchiMate 3.1: "A business service represents explicitly defined behavior that a business [role, actor, or collaboration] exposes to its environment."


The definition in this chapter is compatible with TOGAF, ArchiMate and other sources 

Service (1) the external or declarative view of an activity or process that produces a result (output or state change) of use to some external actor(s).


A service contract typically contains:


·       Name (e.g. build a house, or book an appointment).

·       Entry conditions (inputs and other preconditions)

·       Exit conditions (outputs and other post conditions, inc, perhaps "value delivered")

·       Qualities of service (speed, volume and other measures, inc. perhaps price).


In EA?

Architects start from the mission and vision of a business, drivers acting on it, and the goals (ideally SMART ones) declared by business directors. Given a system of interest, they should identify at least the primary services provided to mee the aims of sponsors and stakeholders. They may simply name the services or specify them in terms of contracts (business and application service contracts in TOGAF).


Beware the term service is used with many other meanings.


Defining how actors are “organized”

Many social entity thinkers address the structures in which actors are “connected”, “organized” and managed”. This section is about what people call a management structure.


Every management structure is shaped by the nature of the business. Armies, coal mines, motor car manufacturers, banks, charities, hospitals and software development projects are likely to have different management structures. However, there are some universal themes in management structure design; many related to the contrasts in this table.


Decentralized or centralized management?

Network connectivity

Hierarchical connectivity

Point to point communication

Hub and spoke communication

Peer to peer cooperation

Client sever cooperation

Self-organization of actions*

Direction and coordination of actions


The term self-organization has several different meanings and possible implications, as discussed in chapter 3b.

Decentralized / network management structure?

Many thinkers see systems thinking as favoring the concepts to the left in the table above. To reduce centralization and hierarchical organization, they recommend flattening hierarchies, cutting out middle managers, and eliminating well-nigh all forms of centralization.


At the extreme, for example in this video, the vision is of a collective, a distributed network of actors or agents, with no central legal entity or bank account, in which no individual has to make decisions or take responsibility.


Outside of informal social clubs, the nice-sounding idea of a collaborative network runs into difficulties. How to bound the network - who decides who the members are? What if members play roles in multiple collectives with conflicting or competing aims?  For a collective to succeed in meeting some aims, its actors need to make decisions and coordinate actions. They must take responsibilities and be rewarded for them. They must be motivated and managed to act in the performance of more or less regular business operations (be they road mending, heart surgery or software development).


It seems thinkers who favor decentralization also tend to favor the concepts to the left in the table below.


Freedom or control?

Ad hoc interactions

Following processes

Bottom-up creativity

Top-down control

Local variation







For most people, the concept of “organization” implies the concepts to the right. If self-organization is characterized by the concepts on the left, then (as Ashby suggested) that seems a confusing contradiction in terms.


Who is the self in self-organization? This graphic captures some ideas found in systems thinking discussion.



Social entity


General principles

Specific aims




Lessons learned


The “self” could be every actor employed in a social entity. Think perhaps of a small charity in which the work is seen as an iterative and incremental learning process, with continual feedback from outputs to inputs in the graphic above.


In a different interpretation, the “self” could be a committee or governing body elected by employees, which determines roles and rules to followed, directs employees to follow them, and may have some power to ensure compliance to the rules.


In most businesses, complete self-determination (whether by employees or an elected committee) is impractical. Most businesses deploy some human activity and socio-technical systems that are designed by people other than the actors who play roles in them.

Centralized / hierarchical management structure?

History tells us that well-nigh all human organizations from the Roman empire and feudal societies to Facebook and the Chinese government today, have a chain of command by which people at the top (tribal chiefs, monarchs, lords, generals, presidents, owners, directors) can direct subordinates’ activities.


In recent times, Google tried eliminating all middle managers, and had to reintroduce some. Some point to open-source software organizations, who advance their product by accepting contributions from distributed, autonomous (sometimes unpaid) contributors. However, there is always a person or small group at the center who determine and control product release; and some open-source project leaders are very autocratic!


Can you think of a business that is a wholly distributed collective of independent agents? In practice, businesses have some directors who impose a hierarchical management structure over their network of employees or volunteers. And even in a small social club or cooperative, it is normal for members to elect some kind of governing committee.


Remember, the systems of interest are not merely passive structures that organize or connect things - be it a classification hierarchy, a matrix, a network, a timetable, or a management structure as shown in an organization chart. Rather, a system is found in the interactions between things, in how actors interact in the performance of activities.


The structure of a human social entity is one thing; how its actors interact is another. In practice, there are many variables in how people behave within a hierarchical structure, including the three in this table.


Interaction variable

Management style

The extent or degree to which


and control


directors use their ability to direct activities



the activities of subordinates are directed in practice



the decisions of directors are informed by feedback from subordinates (and shareholders)




So, the existence of a hierarchical "chain of command" does not necessary imply a command-and-control management style. It remains possible that actors are free to determine most of the actions they perform, and contribute to director's decision making.


In any case, even when working under a hierarchical management structure, actors usually interact in a network to complete business processes (be it road mending, heart surgery or software development) by peer-to-peer cooperation.



In a business, complete self-organization is impractical. To a greater or lesser extent, actors must be motivated and managed to act as the business directors wish them to act. An organization's management structure is usually a compromise between many possible structures - each having its own pros and cons.


And although human actors are autonomous entities, free to do what they choose; they often choose to act in ways constrained by roles in regular processes. Else they would be unable to mend roads, do heart surgery or deliver working software.


In short, the challenge in practice is not to choose between distribution or centralization, freedom or control, it is to get the benefits of both by finding the sweet spot between the extremes.


Other hierarchies in enterprise architecture

In Meadows’ description of a system, the actors are the most concrete and tangible elements, the activities are harder to see, and the aims are even harder to see. In EA, business architects often try to actors, activities and aims visible by drawing composition/delegation hierarchies of the kinds below.


·       An actor hierarchy – as in a management structure.

·       An activity/ability hierarchy - a functional decomposition or capability map.

·       An aim hierarchy – this decomposes grand aims or goals into finer-grained objectives.


Sometimes one hierarchy corresponds closely to another. Sometimes the three hierarchies differ.

Constructing an aim hierarchy

See earlier discussion.

Constructing an actor hierarchy

This section discusses how actors may be composed into a social structure, from small to large.

Grouping actors into productive teams

Wherever actors must communicate and cooperate to do something, it seems the size of a productive team has not changed. It is still around the 6 to 8 foragers who can support a moderately-sized tribe of people. A basketball team needs 5, baseball 9, and soccer 11. The optimal size for a board of directors, a software development team, or a meeting of any kind, is often said to be 6 to 9 people.


 "According to Wittenberg, while the research on optimal team numbers is “not conclusive, it does tend to fall into the 5 to 12 range, though some say 5 to 9 is best.” This source.


In business, how productive teams form and reform is discussed in this source.


"Dynamic Reteaming is team change. Teams evolve and change structurally according to 5 base patterns: one by one, grow and split, isolation, merging and switching. These changes happen at different levels: individual, team, team of teams, department, company, industry and beyond." This source.


How about scaling up from a team to a wider community or organization?

Grouping actors into hierarchical organization structures

The designers of organization's management structure should consider its role in four kinds of relationship.


In a delegation/reporting hierarchy, work is divided between divisions, departments and teams. Usually, information about the progress of work is reported upwards, often in a summary form, to higher ups. Note that at any or every level from large to small, organization units may be paired in a network of customer-supplier relationships.


The traditional management hierarchy widens exponentially from the top down. Say 1 > 7 > 49 > 343 > 2401 > 16,807. As a reporting structure this reduces the information load on each manager. But distances higher level managers from bottom-level operations. Hence the advice for business directors to “walk the shop floor”.


Many deprecate hierarchical command and control. They promote distribution of authority and anarchical communication networks. But a hierarchical organization structure does not necessarily imply top-down command and control. It does not tell you or dictate how autonomous the actors are. Whatever the structure, actors may be either largely autonomous agents or else largely incapable of acting without direction or cooperation.


Perhaps the earliest large human organizations were armies. Most armies are organized hierarchically. But even army commanders delegate some authority, down to the platoon level. And most are trained to adapt to feedback from below. As most famously advised by Helmuth von Moltke (1800-1891).


In a competence hierarchy, the higher ups assist those below them. This assistance can take the form called “the servant leader”, where the role of the manager is to make life easier for those who report to him/her, finding resources, clearing away obstacles.


The trick is to get the benefits of delegation, reporting and competence hierarchies, without the de-motivating effects of the next.


In a dominance hierarchy, the higher ups impose their will.


In many animal communities, from chickens to primates, there is a dominance hierarchy. "Social hierarchy is a key element in the organization of many human and nonhuman groups that undertake collective and cooperative activities. Dominant group members standing at the top [have] benefits including priority access to limited resources (food, mates, space). To maintain their position, dominant individuals exercise behavioral tactics that rely on subtle interplay between cooperation, affiliation, and aggression directed toward subordinates. In contrast, submission, though it limits the access to resources, allows weaker and less skillful individuals to minimize consequences of aggressive encounters [they don't get beaten up]." Quoted from this paper.


Neuro-scientific studies conducted during the last decade have pointed to the role of the serotoninergic system in establishing social hierarchies in primates and humans.

"In macaques, alpha males have twice the level of serotonin in the brain as subordinate males and females. Dominance status and CSF serotonin levels appear to be positively correlated." Quoted from Wikipedia on seratonin.


In business, a decision is often made by the socially dominant members of a team (even when a matrix comparing options may be used to rationalize the decision). Sometimes that is OK, because the dominance hierarchy reflects a competence hierarchy, but not always.

Constructing an activity or ability hierarchy

In EA, a major concern is the cross-organizational standardization and integration of business activities that are, or should be, digitized – regardless of the management structure.


It is normal to draw a logical function or capability hierarchy. This imposes a composition structure of business activities or abilities, to help people make sense of a business.  It decomposes broad-ranging activities or abilities into finer-grained one. It gives architects and other stakeholders a logical and stable picture of what a business does or should be able to do.


Aside: the theory is simple; the practice is not so simple. Different stakeholders may have different ideas of which activities belong together. They value clustering activities using different "cohesion criteria". So, there is no function/capability hierarchy all are happy with. There are as many hierarchies as there are different, valued, cohesion criteria.


I recall (c20 years ago) the CIO of a government agency describing their practice. He said the different interests of 9 board members were represented in 9 different hierarchies. They settled on one primary hierarchy for most purposes, but still, the forest of trees had to be considered.


EA maps the chosen logical hierarchy to the real-world or physical organization structure that is or will be imposed over the human actors. They may do this either perfectly (one to one) or in a matrix structure. When the mapping is one function to one organization unit this creates a so-called" functional organization structure".


Often, business directors and managers divide the real-world organization structure in other ways, for more or less good reasons. They may divide by location, product type, customer type, or a combination of such cohesion criteria.


On the use of organization, function and capability hierarches in EA, read this slide show.


The top level of a function/capability hierarchy may be shown in a value chain diagram. This imposes a structure on the activities of a business. It gives the impression of the business as a single coherent activity system. But it is such a superficial view that it likely masks disco-ordinations between activity systems (e.g. billing and customer service) at a lower level.

People factors

This is a book about terms and concepts used when discussing and modelling systems that operate in the real world. The book does not set out to undermine the whole of “management science” or “systems thinking”, but it does set out to unscramble and debunk some ideas presented under those headings.


This final section adds some remarks on people factors, based on my experience of working in organizations large and small.

On personality classifications

Managing people is difficult. Managers are drawn to schemes that promise to simplify the choices they have to make about how people will be organised and what roles they play. And among the worst sins of business management is the abuse of personality classification schemes


This is an area in which scientists struggle to keep up with the pseudo-science. For example, read the entries in Wikipedia on.

·       "Neuro-Linguistic Programming" and the

·       Myers-Briggs (personality) Type Indicator

·       The HEXACO model,

·       Belbin’s classification of team roles (not personality types)

·       Eysenck's model of personality types

·       Jungs’ Enneagram of personality types.


Even the more scientific of those schemes are of little practical use in management science.

“To say that you are a this or a that, that’s a mistake,” says William Revelle…Robins cautions there is a risk of “arbitrarily drawing a circle around a particular cluster of people…Revelle agrees. “Breaking it down into one of four types would not allow me to understand you very well,” he says. “If I want to know what you’re like, I need to know how able you are to do something, how stable you are, how interested you are in things. I need to know all of that to predict or understand what you’re going to do.”


As Dr. Kevin Murphy, Pennsylvania State University has stated: “The problem with personality tests is [their] validity as predictors of job performance is often disappointingly low. The argument for using personality tests to predict performance does not strike me as convincing in the first place.”.


The validity of using these tools is questionable not only because their predictive value is weak, but also because classifying people becomes a self-fulfilling prophecy. Once somebody has been pigeon-holed as one of a type, managers may stop making the effort to understand and manage the individual. Instead, they start treating the person as one of the type. Every conformance to type is used as evidence to confirm the classification, every deviation from the type is disregarded as exceptional.


By managing a person as one of a type, managers force the person to fit the type (or to rebel and be seen as a trouble maker). Thus, a manager comes to disregard a person’s flexibility or ability to occupy other pigeon holes. In short, be very wary of pigeon-holing people using personality classification schemes. See also the “Human Factors in Hierarchical Organisations” page at       

On “culture”

It is one thing to believe and agree the culture of a business is important, for any of several possible reasons. It is another to pin down what that means in scientific way.


"Culture” is defined by The American Sociological Association as “the languages, customs, beliefs, rules, arts, knowledge, and collective identities and memories developed by members of all social groups that make their social environments meaningful."


Social animals have evolved the ability to verify/falsify the meanings of memories and messages by interacting with things, including other animals, in the world. That is a psychobiological fact of life. Beyond that, the definition of culture is so variegated and all-embracing it is impossible to define the culture of any substantial social group, or use it predict anything. The definition has some challenging implications.


A social group is bounded within an environment? Then how to bound it? Who defines its members? We no longer identify with one tribe; we identify with any number of groups with different, sometimes conflicting, aims. We divide our time between groups, with differing degrees of commitment to each group.


A social group has one coherent culture? The definition implies every individual member's language, customs, memories (etc.), is part of the culture of every group they belong to. In other words, the culture of any group you observe is an aggregate of its members' individual, sometimes conflicting, cultures. And the culture is continually shifting as a result of people joining and leaving the group, with their ever-changing personal ambitions, hopes and fears, and ever-changing interpersonal relationships.


A social group develops a culture? The reverse may be true. We form groups by identifying and interacting with people who share something of our own language, customs, memories etc. So does a collective develop a shared culture, or does a shared culture lead people to form a collective?

The servant leader

The idea of a confident leader setting goals and accepting inputs from below dates from the 19th century if not before. Try Helmuth von Moltke (1800-1891). Today, it is commonly said that organization or project managers should be "servant leaders". The idea is good, it can work well, but in practice is often difficult to apply. It is easy and natural to be a servant leader when a) those you lead know as well or better than you and b) you are confident in your position as leader. Very often, one of those conditions is not fulfilled, and the result can be a dysfunctional organization.

Causes and symptoms of dysfunctionality

As I see it, the root causes of demotivation and dysfunctionality usually lie at the personal and interpersonal level. And in the natural desire of people to maintain or advance their own role in an organization. The more trusting "culture" that we'd all like to work in has to address problems that individuals create by one or more of:


1.      Unclear, ambiguous, communication

2.      Fear of exposing ignorance

3.      Fear of sharing knowledge

4.      Not appreciating, praising, thanking, rewarding

5.      Acting to keeping others apart, prevent communication

6.      Refusing to hear or accept criticism or alternative views

7.      Blaming and blame passing

8.      Absenteeism - physical or virtual

9.      Incompetence

10.   Animosity, unfriendliness, unkindness.


Minimizing issues arising is a task for everybody, and managers in particular.


In a business, the recruitment, motivation and management of human actors to play roles is primarily the concern of business managers, rather than EA. This chapter says nothing about incentivizing human actors, by means of salary, bonuses and other benefits, and events designed to improve esprit du corps.


Wrt overlapping social entities: You no longer belong only to one family and small tribe. In the modern "information age", your memberships of social networks are fluid. You can identify with any group you like. You can celebrate you belong to infinite nameable groups. Best not to let yourself be defined by any of them, or get sucked into "identity politics".


Wrt diversity: If men and women differ, we can expect them to choose different roles. If they are the same, we can expect roles to be divided between the sexes, but to imply one sex should learn from the other would be patronizing or matronizing.

Conclusions and remarks

This chapter addresses some points relevant to the success of business entities, notably principles to be followed, aims to be achieved, and structures for organizing interconnected human actors. It does not attempt to introduce more than a small fraction of the ideas to do with organization design and business management.


One take-away: much naive systems thinking (some pseudo-scientifically referring to "graph theory") promotes distribution, network, point-to-point communication and self-determination, and deprecates centralization, hierarchy, hub-and-spoke communication, and directed action. Structure is one thing; behavior is another; and all is trade offs.


Social entity thinking focuses on how actors are directed, motivated and organised to produce outcomes; on defining general principles for actors to follow, specific aims for activity and not only the structure under which actors managed but how that structure is used.


In EA?

Although EA primarily strives to extend, improve and coordinate activity systems, some social entity thinking is needed as well.


For some more sociological views of society



The adaptability of an enterprise depends on the extent to which a) actors are able and free to act creatively in ways they judge best for the business and b) activity systems are “designed for change” ahead of time to include actions that actors could conceivably be asked to perform.



Appendix: Some famous sets of principles

Kelly’s 14 rules for agile airplane design and build

1.      The Skunk Works® manager must be delegated practically complete control of his program in all aspects. He should report to a division president or higher.

2.      Strong but small project offices must be provided both by the military and industry.

3.      The number of people having any connection with the project must be restricted in an almost vicious manner. Use a small number of good people (10% to 25% compared to the so-called normal systems).

4.      A very simple drawing and drawing release system with great flexibility for making changes must be provided.

5.      There must be a minimum number of reports required, but important work must be recorded thoroughly.

6.      There must be a monthly cost review covering not only what has been spent and committed but also projected costs to the conclusion of the program.

7.      The contractor must be delegated and must assume more than normal responsibility to get good vendor bids for subcontract on the project. Commercial bid procedures are very often better than military ones.

8.      The inspection system as currently used by the Skunk Works, which has been approved by both the Air Force and Navy, meets the intent of existing military requirements and should be used on new projects. Push more basic inspection responsibility back to subcontractors and vendors. Don't duplicate so much inspection.

9.      The contractor must be delegated the authority to test his final product in flight. He can and must test it in the initial stages. If he doesn't, he rapidly loses his competency to design other vehicles.

10.   The specifications applying to the hardware must be agreed to well in advance of contracting. The Skunk Works practice of having a specification section stating clearly which important military specification items will not knowingly be complied with and reasons therefore is highly recommended.

11.   Funding a program must be timely so that the contractor doesn't have to keep running to the bank to support government projects.

12.   There must be mutual trust between the military project organization and the contractor, the very close cooperation and liaison on a day-to-day basis. This cuts down misunderstanding and correspondence to an absolute minimum.

13.   Access by outsiders to the project and its personnel must be strictly controlled by appropriate security measures.

14.   Because only a few people will be used in engineering and most other areas, ways must be provided to reward good performance by pay not based on the number of personnel supervised.

Agile software development principles

As in the agile manifesto.

1.      Our highest priority is to satisfy the customer through early and continuous delivery of valuable software.

2.      Welcome changing requirements, even late in development. Agile processes harness change for the customer's competitive advantage.

3.      Deliver working software frequently, from a couple of weeks to a couple of months, with a preference to the shorter timescale.

4.      Business people and developers must work together daily throughout the project.

5.      Build projects around motivated individuals. Give them the environment and support they need, and trust them to get the job done.

6.      The most efficient and effective method of conveying information to and within a development team is face-to-face conversation.

7.      Working software is the primary measure of progress.

8.      Agile processes promote sustainable development. The sponsors, developers, and users should be able to maintain a constant pace indefinitely.

9.      Continuous attention to technical excellence and good design enhances agility.

10.   Simplicity--the art of maximizing the amount of work not done--is essential.

11.   The best architectures, requirements, and designs emerge from self-organizing teams.

12.   At regular intervals, the team reflects on how to become more effective, then tunes and adjusts its behavior accordingly.

Deming’s business management principles for quality

1.      Create constancy of purpose toward improvement of product and service, with the aim to become competitive and to stay in business, and to provide jobs.

2.      Adopt the new philosophy. We are in a new economic age. Western management must awaken to the challenge, must learn their responsibilities, and take on leadership for change.

3.      Cease dependence on inspection to achieve quality. Eliminate the need for inspection on a mass basis by building quality into the product in the first place.

4.      End the practice of awarding business on the basis of price tag. Instead, minimise total cost. Move towards a single supplier for any one item, on a long-term relationship of loyalty and trust.

5.      Improve constantly and forever the system of production and service, to improve quality and productivity, and thus constantly decrease costs.

6.      Institute training on the job.

7.      Institute leadership. The aim of supervision should be to help people and machines and gadgets to do a better job. Supervision of management is in need of an overhaul, as well as supervision of production workers.

8.      Drive out fear, so that everyone may work effectively for the company.

9.      Break down barriers between departments. People in research, design, sales, and production must work as a team, to foresee problems of production and in use that may be encountered with the product or service.

10.   Eliminate slogans, exhortations, and targets for the workforce asking for zero defects and new levels of productivity. Such exhortations only create adversarial relationships, as the bulk of the causes of low quality and low productivity belong to the system and thus lie beyond the power of the work force.

11.   Eliminate work standards (quotas) on the factory floor. Substitute leadership. Eliminate management by objective. Eliminate management by numbers, numerical goals. Substitute leadership.

12.   Remove barriers that rob the hourly paid worker of his right to pride in workmanship. The responsibility of supervisors must be changed from sheer numbers to quality. Remove barriers that rob people in management and engineering of their right to pride in workmanship. This means, inter alia, abolishment of the annual or merit rating and management by objective.

13.   Institute a vigorous program of education and self-improvement.

14.   Put everybody in the company to work to accomplish the transformation. The transformation is everybody's job.