Typology is a great tool for learning and using theories. We can consider typology as a type of intermediate knowledge for connecting theoretical research and domain practice. As an information architecture, I create typologies for curating data and information, designing websites and apps’ structure. I also collect typologies for learning new knowledge.

In 2014, I started learning Activity Theory. After reading some books and papers about Activity Theory, I moved to search for some typologies of activities. In 2015, I found Clay Spinuzzi’s paper Toward a Typology of Activities: Understanding Internal Contradictions in Multiperspectival Activities. He pointed out, “AT (Activity Theory) currently lacks a suitable typology for characterizing ideal types of activities in terms of multiperspectivity, so it has had trouble systematically characterizing the resulting sets of internal contradictions.”

Spinuzzi (2015) summarized several typologies made by other activity theorists before presenting his own idea. For example, “Y. Engeström (1987) characterized activities in terms of craft activity, rationalized activity, humanized activity, and collectively and expansively mastered activity…In his later work, Y. Engeström (2008b, pp. 190–191) drew on Victor and Boynton (1998) to describe historically different types of production (craft, mass production, lean production, mass customization, and coconfiguration), each with its own objects and contradictions.” Spinuzzi commented on Engeström’s typology and pointed out its limitation, “…this typology depicts a historical progression of separate ideal types of activity with separate axes, not a matrix with the same axes — making comparison difficult. ”

Other activity theorists created typologies of activity with a 2x2 matrix format. For example, “…Engeström, Brown, Christopher, and Gregory (1997) used a matrix with the axes of flexibility and collectivity to characterize a zone of proximal development whose quadrants represented professional craftwork, market-driven case management, bureaucratically regulated work, and informal, networked processing…Jarzabkowski (2003) proffered a typology of activities based on the axes of actors, collective structures, and strategic activity…” Spinuzzi argued that “these matrix-based typologies have tended to not be broadly applied, serving rather as a way to characterize specific cases. More important, they have tended to not illuminate internal contradictions based on competing perspectives within the activity.”

Spinuzzi came to the conclusion that there is a need to make an ideal typology of activities for general purpose to “explore internal contradictions resulting from hybrid activities”.

How did he create an innovative idea? The answer was starting from the object of activity.

It is easy to create a 2x2 matrix, however it is hard to build a great matrix for general purpose. The key is to find the right object and right dimensions for understanding the object. Now we can claim that Spinuzzi’s choice is a smart decision.

How and Where

Spinuzzi said, “To develop an adequate typology of activities, then, we must characterize these objects — keeping in mind that since objects are multiperspectival, specific activities will often appear as hybrids located within this typology rather than fit neatly into a given type.”

He proposed a new matrix with the following two dimensions:

  • How is the object defined? Is it defined explicitly and deductively or tacitly and inductively?
  • Where is the object defined? Is it defined within the activity’s division of labor or outside it?

The outcome is a matrix with four quadrants (see the diagram below) which represent four ideal activity types.

Spinuzzi named these four ideal types of activities as Hierarchies, Markets, Clans, and Networks. We have to notice these words are just names of ideal types. Spinuzzi said, “I use the term markets in the sense of most organizational typologies (e.g.,Adler & Heckscher, 2007; Boisot & Child, 1999; Boisot et al., 2001; Cameron & Quinn, 2011; Ouchi, 1980; Ronfeldt, 2007), not strictly in the sense of commerce. Clan may be too precise; I prefer Adler and Heckscher’s (2007) ‘community.’ But the term community already denotes a different concept in activity theory.”

Activity theorists see an activity as a system. We have learned from Engeström’s Activity System model, an activity system includes several elements such as Object, Subject, Instruments, Outcome, Rules, Community, and Division of labor. Following this model, Spinuzzi discussed four ideal types of activities.


  • Object and outcome: Hierarchies excel at pulsing explicitly and internally
    defined objects, producing highly predictable outcomes.
  • Tools and rules: Hierarchies tend to use well-defined tools and rules to keep the pulse steady, the object explicitly defined, and the outcome predictable. Information tools such as databases tend to demand structured, explicit information. Knowledge about this information tends to be explicit and formalized rather than tacit.
  • Actors and community stakeholders: Hierarchies require a great deal of internal trust; they are trust intensive. They tend to focus internally and to define the outcome in terms determined by the actors who pulse the object, not by the community stakeholders who receive it.
  • Division of labor: Hierarchies demand control. So the division of labor involves clear lines of authority. Relationships between actors tend to be dependent — with those lower in the hierarchy depending on those above them — and trust intensive.


  • Object and outcome: Markets are good at pulsing objects that are explicitly defined and differentiated — objects that compete with other objects. The outcomes tend to be externally defined; an organization must give up control to seek market solutions. But in return, the organization gains the outcomes of competitiveness and flexibility.
  • Tools and rules: For market solutions to work, inputs and outputs have to be explicitly specified and often standardized. Tools must include highly structured communication, such as exchange prices — a high level of abstraction and codification — so that the market can perform the job competently and efficiently without too much supervision. Rules are spelled out in formal contracts and enforced via reciprocal self-regulation: The market determines the price.
  • Actors and community stakeholders: In a market, the object is defined…by people in other activities who purchase it. Relationships between actors and community stakeholders are not necessarily steady from exchange to exchange, pulse to pulse.
  • Division of labor: Relations between activities in a market are independent and formal, based on the exchange of goods and services. The labor is coordinated horizontally via self-regulation; that is, buyers and sellers coordinate their negotiations via highly structured but contextually shallow information such as price and specifications.


  • Object and outcome: Clans work well for addressing customized objects, such as craft objects. The objects themselves are defined tacitly rather than explicitly and often inductively rather than deductively. They are also internally defined, typically leading to outcomes of team identity or belonging. Outcomes are often customized. In the case of internal culture, clans inductively develop their own internal values over time, based on their internal needs and experiences.
  • Tools and rules: Since objects are defined tacitly and internally, they often involve tools that are unstructured and diffuse. Face-to-face coordination is common, especially in terms of mutual adjustment. In this context, rules tend to be commonly held traditions, values, and beliefs within the tight-knit group of actors.
  • Actors and stakeholders: Actors are especially tight knit as they pulse the object. They sometimes become very clannish, defining themselves against others in the same organization.
  • Division of labor: Since they are so tight knit, clans tend to have a very flexible division of labor, coordinating horizontally via negotiation. They are interdependent, that is, dependent on each other as they coordinate to pulse the object.


  • Object and outcome: Networks pulse objects that require various types of expertise, particularly objects that require collaboration across specialties (and thus activities)…This configuration is especially well positioned for producing innovative outcomes, outcomes that cannot be achieved by individual specialties.
  • Tools and rules: The tools used by networks can be, and often are, unstructured, diffuse, and loosely coupled. They are pulled together across specialties to attack unique problems, so they are often cobbled together for a unique engagement. Rules are emergent and tend to develop over the life of the collaboration.
  • Actors and community stakeholders: Actors in networks can be internal to an organization or cross-organizational. Like actors in clans, actors in networks tend to be interdependent: Specialists need each other’s contributions if they are to complete the project.
  • Division of labor: Networks do not have a necessary center of control; typically, they are horizontally controlled via the emerging collaboration. They tend to establish interdependent relations, and they must develop swift trust in order to work well.

Spinuzzi also pointed out, “…organizational typologies generally acknowledge that organizations tend to be hybrids, not ideal types. Similarly, we can understand.” He also noticed that “…at different points, the same object can be understood as the object of a hierarchy, a market, and a network.”