If it’s not AE, it’s not EA

What is enterprise-architecture?

According to just about everything you’ll see at present, from books to so-called ‘standards’ to conferences to recruiters’ job-adverts, it’s kinda something to do with IT, sorta, ish… uh, kinda dunno, really, but it’s sort-of IT and it’s sort-of enterprise and it’s sort-of architecture, innit…?


Well, there’s a really simple test we could do – in English, at any rate – and that’s to swap the words around and see if it still makes sense.

Hence data-architecture is the architecture of data.

Applications-architecture is the architecture of applications.

Infrastructure-architecture is the architecture of infrastructure.

Security-architecture is the architecture of security.

Brand-architecture is the architecture of brands.

And enterprise-architecture, we’re told, is the architecture of IT.


This – to quote a certain well-worn phrase – does not compute.

Let’s perhaps do this a bit more systematically?

Enterprise is a human construct. It’s about people – about people being enterprising. When we (mis)use the term ‘the enterprise’ to mean ‘the organisation’, it’s actually about a specific type of context within which people can perhaps be enterprising. Technology can sometimes help enable people to be enterprising, but in no way is the technology itself ‘the enterprise’.

And architecture is not just about structure, but again, primarily about people – about where people’s intent and, yes, enterprise intersect with structure and story. In that sense, architecture is as much about story as it is about structure.

The architecture of the enterprise is about the intersection of structure and story to support people in enterprise.

If it’s not primarily about people, it can’t be about enterprise.

If it’s not primarily about the intersection of structure and story, it’s not architecture.

In short, if it’s just about the structure of a specific subset of technology – which is pretty much all that most people still seem to mean by their (mis)usage of the term ‘enterprise-architecture’ – then it’s neither enterprise nor architecture. And hence a dangerously-misleading misuse of an otherwise very specific term. A misuse that actively misleads everyone in the entire enterprise-architecture space, and actively prevents those who do work with ‘the architecture of the enterprise’ from being able to describe what it is that they do.

So here’s a really simple test to check whether you’re doing enterprise-architecture, or merely doing large-ish scope IT: if it’s not the architecture of the enterprise, it’s not enterprise-architecture.

If it’s not AE, it’s not EA.

‘Nuff said…

5 Comments on “If it’s not AE, it’s not EA

  1. I’m reminded of what Steve Jobs said:

    “It’s in Apple’s DNA that technology alone is not enough — it’s technology married with liberal arts, married with the humanities, that yields us the result that makes our heart sing…”

    And I heard a rumour that Apple are quite successful as a company, as an enterprise.

  2. There is a time to analyse and a time to synthesise.

    If we find ourselves constrained by being on the wrong side of a dichotomy between ‘IT’ and the ‘Enterprise’ one strategy is to relocate to the other side.

    Another strategy is to challenge the dichotomy itself.

    As I am sure you have intuited Tom, I am disinclined to join the migration to the Enterprise. Not because I am an advocate for EITA. I am not.

    I am just an old Heideggerian in EA’s clothing. I think our being is in our use of tools. An in that I am quite aligned to your turn to narrative. But what we have at our disposal to use is a determining structure of the stories we tell.

    I am thinking now of your story about the folding bike. The bike, the trolleys, the checkout, the store itself, the roads you cycle on – all are technology, some things in the story, visible or not, would have been information technology.

    The story – the conflict of intentionality that is captured – is structured by how the actor’s intentions are in turn structured and mediated by the technologies into which those intentions are projected.

    In today’s world every story is mediated at some point by technology. Organisations – Enterprises – are often expressed in their ‘IT’ systems.

    I remember seeing a very bright person I know present on EA – and he has set up a quite successful practice in a large enterprise. He addressed the issue of whether EA should be in ‘IT’ or ‘the business’. He said, “I don’t care, but whichever side it is on, it needs to be by the border.”

    I liked that.

    When it comes to EA and the Enterprise or IT, I would first ask whether the either/or assumption is valid.

    For now I see nothing lacking in both/and.

    • Hi Ric

      We’re not disagreeing, and to me at least there’s no either/or – that’s something you’ve assumed about me, but not something I hold at all.

      (You’re not the only one, by the way. Some people were shocked the other day when I asked for more detail about the IT in an architecture, they said “But you’re the anti-IT guy!” – to which I replied “I’m not ‘anti-IT’, I’m the ‘pro-wholes’ guy – if the IT is missing from a spec, I want to know about it!”)

      For the architecture of the enterprise, there are two issues in IT-centrism, or anything-centrism in general: one is that the scope is too narrow, the other is that centring the attention around anysingle ‘the centre’ will lead to a dysfunctional architecture at the whole-of-architecture scope. In a viable architecture, everywhere and nowhere is ‘the centre’.

      Remember, too, that the concept of ‘the enterprise’, as something significantly larger than the organisation-in-scope, both provides the organisation-in-scope with its context, and sets the scale and scope for the overall architecture and its influencers. In this at least, I largely agree with the current conceptual-structure of TOGAF (as long as we remember that the actual ‘the organisation-in-scope’ for TOGAF is solely IT-infrastructure, nothing more than that): in TOGAF, its ‘business-architecture’ is, in essence, ‘anything not-IT that might affect IT’, whereas for a whole-of-organisation’s architecture, ‘the shared-enterprise’ is ‘anything not-the-organisation that might affect the-organisation’.

      (Where people so often get TOGAF wrong – spectacularly, even lethally wrong – is in trying to run the same structure upside-down, placing everything-IT as the outer rim of the scope, with everything else inside it… But that’s another story, of course…)

      A viable architecture is always about the-whole-as-whole. We don’t centre on ‘the enterprise’ any more than the IT or the buildings or the finance or security or anything else: there is no ‘side’ on which to ‘take sides’. It’s all one and the same – it’s all the same architecture.

      I do know that you get this point, because it’s implicit in all of your own writing, and it’s not possible to develop a viable architecture without it; it’s just that it’d be kinda nice if you didn’t automatically assume that I don’t get it?

      Bah grump grump grump… 🙂

  3. It is relatively easy to talk about the-whole-as-whole on paper but in reality not many wholes have clear boundaries. E.g. what are the boundaries in space, time and society for the Rijksmuseum* as a whole enterprise? And how much of that whole is under the design influence of the architect(s)?

    Isn’t the whole just a perception (formed by experience) of the architect of what should be under the architect’s influence?

    *see https://www.rijksmuseum.nl/en/renovation/the-renovation

    BTW: it would be interesting to see, with hindsight, if TOGAF/ArchiMate would be helpful in the case of redesigning the Rijksmuseum (not just the building but as an Enterprise) 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *