Heterarchies
Many discussions about change in organizations are about the demise of hierarchies and the rise of the networks. Sure, this is a trend that can be seen, but there are not many organizations without hierarchy, and I don’t think hierarchies will diminish completely. On the contrary, hierarchies have a valid function and purpose, there are familiar and relatively simple. However what we do see, is that organizations become flatter, layers are becoming thinner or even removed, and people connect more with other people by means of technology.
Karen Stephenson acknowledges this as well, and comes with an interesting point of view: heterarchies (PDF link to article). The heterarchy consists of at least three separate hierarchies that have their own responsibilities, but must collaborate to achieve a collective good that is too complex to achieve on their own. She defines the heterarchy as follows:
A heterarchy is an organizational form somewhere between hierarchy and network that provides horizontal links permitting different elements of an organization to cooperate, while they individually optimize different success criteria.
What she seems to say, is that hierarchies have their disadvantages that are removed by networks, but either the latter doesn’t work in reality or is too complex. She’s seems to search for something is between, the best of both worlds.
According to Stephenson, it is important to have these different hierarchies engaged. Key is collaboration instead of competition. Partnerships between organizations as you wish, or between business units within large corporations. And she admits that this is not easy at all. When you try to map a large organization as a heterarchy, you have to find connectors. The table below compares the market, hierarchy, network en heterarchy on some features. It focuses on its strengths.
I am not looking for a proper definition of heterarchies, or whether you agree with Stephenson or not (well, I’m curious for that of course), but I am more interested in how you can identify people or hubs in an organization that is a connector to other parts of the organization, but not in a hierarchical way. This identification can make such organizational forms less complex. But how do you map these people? Are they certain types of people, who you can trust? Do they have to have certain positions in an organization? Stephenson suggests the following steps:
- Send out a survey where people identify other people that you think are innovative, have integrity, work hard to achieve goals, that you depend upon, and ask people who should be surveyed as well.
- Find connectors by means of interviews. People that score high on the surveys can be persons to ask questions to validate the survey.
- Connect connectors so they can exchange information, knowing that they need each other. They can connect organizational silo’s and collaborate instead of compete.
According to Stephenson there are three types of connectors, or actors in these heterarchies, hubs, gatekeepers and pulstakers. Hubs know a lot of people and act as facilitators, gatekeepers are critical connections between networks and help people to focus, and pulstakers are asked for their opinions and guard the integrity. So if you can map an organization more like a network, or like Stephenson, as a heterarchy (I’d rather call it the informal connections), what’s next? How can these hubs or connectors be more of use to the organization, how can their strengths be utilized better?
Bas Reus' quest on self-organization and online collaborative spaces 
Interesting post, Bas. My first thought is to imagine the possibilities if all organizations’ people could move to working in a heterarchical way or if they would simply “allow” their people to leverage the informal connections, as you refer to them.
Kathi, yes I agree with you on that. Allowing people to leverage informal connections is something an organization can thrive for. But the question remains, how can one do that? What possibilities come to mind when you try to imagine these? Would you first try to identify these connectors, and if you would, what is the next step?
I would probably start by acknowledge their roles in the informal network, and try to give more tools to them so they can fulfill their role as a connector easier. And I would probably try to arrange meetings for connectors to meet with other connectors, ideally so that connectors will then find other connectors themselves afterwards. What are your thoughts on that? What influence could one exercise?
I think the idea of conceptualizing ideas such as heterarchy, hierarchy, networks etc can be of value but more as thinking tools rather than actual things. For me the key difference when we think of these things is the way that we imagine, or experience the way power plays a role in the interactions between people in these various concepts of organizational design.
There would be numerous example of organizations designed with a very strong hierarchy that acted as a network due to the power dynamics that informed the interactions between people. And I think some of the problems with the complexity of collaboration technology have to do with it being asked to overcome constraining power dynamics that have little to do with the technology itself.
We are likely better served to discuss the power dynamics currently at play in the interactions between people and work to change those; I would imagine a new design would follow this and collaboration technology much less complex.
As for finding the connectors, this is an interesting concept I think and one I grapple with. Many of those connectors are people that are in the background of the organization and it is this fact that helps them play this role. Once identified, they typically are no longer in the background and thus their role as a connector changes. The very identification of them can eliminate their unique value and you have to go looking for the new connectors. I think sometimes it is best to try and identify the ‘how’ of these connectors more than the ‘who’ of them and perhaps try out the how in a broader context.
Again an interesting post on your blog Bas. And a very cool pdf: it is nice to see the same subject seen from different perspectives. It also makes clear that people agree on the “should we”, but there is a lot unclear on the “how”.
In my opinion allowing people (like Kathi implied), and enabling people (tools), is indeed important, but is it more important to get the incentives right: Are all the parties that should be involved in a heterarchy convinced of their gain? If they are, I think they will find the ways and means, I do not think this will succeed the other way around.
@tmsamericas (Tom), I think you’re identifying an important issue here.
“Many of those connectors are people that are in the background of the organization and it is this fact that helps them play this role. Once identified, they typically are no longer in the background and thus their role as a connector changes.”
You could be right here. What happens when you formalize an informal role? What changes then for connectors and their connected nodes? I think identification can be valuable, but perhaps without formalization. Then they can act the way they always acted, without experiencing new responsibilities.
@kraker: thanks for the kudo’s. Your note complements that of Tom I think. Allowing and enabling is important, but the effect of formalizing can be undesired. My question, should all parties (or the connectors) be involved, or should they just find their tools themselves, because their availability is better? In other words, should you make them known with their role, or should they just continue what they do without knowing their role as connector?
Lots of great questions – one leading to the other: Include the connectors? Identify the connectors? What role does power play? From my experience, the right anwsers to all of these, and the right answer will differ org to org, comes down to trust of those you work with, those in your network/your hierarchy/your informal connections. It is the critical piece for collaboration.
Bas, This is an outstanding (if not breakthrough) analysis of organizational design options. I have been contemplating alternatives to hierarchy for some time, and I think you and Karen have done an excellent job of framing a hybrid “heterarchy” model – including her original paper, and your analysis of it. Adaptive social emergence at its finest.
Re: connectors, I’ve been grappling for a ubiquitous way to identify people in a personal or community/organization network as well.
I have been calling my quest “In search of SMEs” (subject matter experts).
I think the connector problem will be solved in time as social media applications move to embrace more consistent profile information, which includes metadata tagging to make SME’s searchable, sortable, etc. This is already staring outside the firewall, and will eventually occur, I think, inside (the Enterprise 2.0 case). I agree power structures and old-paradigm thinking will impede rapid movement to these new models, but I think truly innovative organizations will realize the collaborative power in heterarchy-style solutions.
If heterarchy is embraced, the role of the connector must be both official and visible.
I am curious: in this thinking, are you not applying a complex adaptive systems (CAS) context for social and commercial organizations? I am a new student in the complexity space, but from my initial reading, I think organizations that can diversify, connect, engage & learn (4 CAS “attributes” per Scott Page perspective; I think of them as “enablers”) will achieve the highest collaboration potential, and in turn, be the most likely to innovate. Back to the SME context, both “diversify” and “engage” dynamics within CAS require awareness of expertise available to the team. Either a connector (as you and Karen define the role) or an online profile database will be needed to make such CAS-optimized capabilities operational, and again, my hope, ubiquitous.
My view of the resultant team: a “learning cell” –
On the above SME profiling challenge, I have some real-world scenarios I can share if you like, including tools that address profiles (and would-be connectors) both inside and outside the firewall, as well as some online Twitter communities working under the umbrella of CAS optimization.
Again, though, this is an exciting line of thinking. I look forward to your additional comments and insights.
sourcepov, thanks for adding to this discussion. Subject Matter Experts (great name) can be of great value if you are searching for expert networks. Of course you can look for other types of networks as well, including a social, innovative or learning network. Other connectors will come to surface there, and can be of value for different purposes.
I’m curious how you look at making the unofficial official and visible, like making the connector role more formal. Would it have impact on the connector? How would it have impact on the network? Are you changing hierarchical structures there?
About the CAS, I must admit that I’m not an insider to this theory. However, if you talk about diversify, connect, engage and learn, I think it depends on the purpose you’re after. Like I mentioned, it can be innovation, it can be learning, something else or a combination. Engaging (e.g. with the outside world) can lead to innovation for example. Connecting can be useful, but with who do you connect, who are connectors and what is it that you are after? I think I have to look into this CAS theory any time soon.
And please share some real-world examples if you like, I think we can all learn from them. It makes me very curious…
As I am new in the complexity space, its hard for me know where CAS stops & starts. I have been studying the work of Scott Page (U.Michigan) and believe he is one of the primary voices on it. I believe it is most heavily embraced by the social sciences & computer-based agent modelers, and is centered on systems that possess an ability to ‘learn’ and ‘adapt’ under appropriate conditions, aka ’emergence’. Systems of people (societies, economies, social ecosystems like ‘education’, communities) are great examples.
Hoping others can shed more light ..
As a student and non-academic practitioner, I focus mostly on applications .. and at an intuitive level, I think I see many.
To pick up on your line of questioning – let’s focus on the objective of ‘learning’. I think this is a short hop to innovation, but let’s keep it at learning.
The goal would be to create a ‘learning cell’ (with our industrial age paradigms, a ‘learning machine’ metaphor might be more intuitive, but for CAS we are best served to keep within an organic paradigm, so ‘learning cell’ is the better model.) As a very quick real-world example of self-organizing systems and the connections that a CAS ‘learning cell’ would need in order to function optimally, check out this new list of Twitter resources:
Complexity & CAS Stakeholder Group: http://bit.ly/casTML
Pardon the recursive example. The subject matter could be anything, basket weaving, blog writing, ecosystem reform. Since you and I are interested in CAS & Complexity, I thought we’d use this one.
So you and I want to learn more on the topic of “CAS & Complexity”. We could create a self-organizing CAS (as ‘learning cell’) that would (a.) establish knowledge of diverse content stakeholders (b.) physically or virtually connect with them .. we’ll stick with virtual for now, it’s cheaper, easier & faster !!, and then (c.) create some means of interdependent activity, which in the social media context, we’d refer to as ‘engagement’; in practical terms, this would take the form of blogs, blog comments, tweets, online chats, SKYPE sessions, emails or phone calls, depending on the degree of 1:1 vs. n:n engagement necessary; in the old, pre-SM world there was a physical proximity constraint, namely: conferences, meetings, institutes, universities.
Before going any deeper on CAS aspect of complexity science, I’d like to get some additional insight from others deeper in the CAS space.
I don’t know that I’m on solid ground w/ ‘learning cells’ just yet ..
Meantime, I can definitely start pulling together some other real world examples of ‘connectors’ that would help drive the heterarchy paradigm. I want to go back through Karen’s paper again (and the responses published with it) to make sure I understand the context she is operating within. If it’s Enterprise 2.0 and the corporate organization, I’ll be on very safe ground w/ my examples ..
Thanks so much for the focused conversation & questions. You have a great knack for asking excellent questions ..
Hej! That’s a fine contribution. Thanks a lot! Looks good to me. In my line of work, I am communicating mainly with e-mail. I preferably work with Outlook as my e-mail client and with the help of Email Sorter Wizard, an Outlook add-on, I get all my email sorted. In your blog one can find tons of valuable info.
[…] we see many terms and concepts that explain different forms of organization. I already mentioned heterarchies, and there are many more that describe networked forms of organization, such as peer-to-peer and […]
IMO Ms. Steohenson’s article tracks closely with our more recent conversation.
By hybrid, I understand that to mean that execs and managers can choose to centralize and decentralize simultaneously, or in parallel, given the (inherent) flexibility of technology enhanced with (for example) hyperlinks and open standards, etc.
Becomes both / and architectural options all the way up .. and down 😉
I like the term “Source Matter Experts” .. others whom you know, who know where to find things you don’t know for which you have interest or need. Did I just coin that term ?
Source Matter Experts, yes nice term 😉
It’s essentially the same as the “innovation dial-tone” that Chris (sourcePOV) mentioned at the wirearchy discussion. I’d like to call it your extended knowledge. You do not have to know everything, instead you need to know who knows it, or knows who knows it. Know where to tap into your extended knowledge.
My understanding of hybrid, is that we have a hierarchy that shapes the organization, and we have decentralized informal connections that is needed for collaboration etc. In the decentralized networks, it is important that you know where to find your information you need, or with which people you can accumulate or gather it. Up and down, left and right, front and behind, everywhere.
Yup.
[…] type of scaling, as Bas Reus shows us, might appear as heterarchy, a wirearchy, or some other networked structure that resembles nothing like the traditional […]
[…] bringen, die formale Macht (alleine …) als nicht mehr produktiv ansehen.Hier bin ich auch auf das spannende Weblog von Bas Reus gestoßen, das mir sicher bei einigen Fragen weiterhelfen […]