RW - Navigation

RW - Recent

Last 10 entries [comments]:


Last 10 posts [threads/views]:


Last 10 pages updated:

There are 487 wiki pages in total.

RSS logoRSS Feed

Ontology - Applying the “Is A” Test

by Nic Plum on Saturday 20 March, 2010 - 20:37 GMT

Posted in Architecture FrameworkArchitecture Modelling


Can we say This is partly triggered by musings on things heard at the Integrated EA 2010 Conference - well worth attending and I very much part of the value in attending is that in listening to presentations and discussions triggers ideas.

There were some common themes including ontology and a running joke or competition throughout the 2 days was seeing which speaker managed to get the highest number of mentions of ‘ontology’ in their presentation. The presentation on Ontologies won and got the prize packet of mints (sweets).

It was inevitable that “systems of systems” got aired. In the definition there was a mention of it being characterised by behaviour that happened that wasn’t part of the design intent. This sounds pretty much like the ‘emergent behaviour’ which is a characteristic of a system.

The “Is a” Test

Lets take the “system of systems” and try a little test to see it it helps. Can we say that “system of systems” is a system? If we can then we should just classify it as such. If not, what is this thing? What does it have that a system doesn’t?

If it isn’t the same as a system how can it legitimately be represented in an architectural using the same stereotype as the system? It can’t. How can you show that it has different attributes or behaviour? You can’t. Conversely if you do choose to use the same modelling constructs then you’re really telling everyone that it is a system after all.

The metamodels for most frameworks of any decency are quite strict about meaning and as we head towards OWL, BORO and things ontological we can’t just invent concepts. If ontology is important as everyone seems to say then we should apply the principles.

I’ve yet to see any argument that shows that “system of systems” is anything but a system. It’s the current version of the old argument about system and sub-system - it just embeds the viewers’ perspective. My system might be your sub-system. Depending on where I draw my system boundary I might have other systems within mine and therefore my system is also a “system of systems”.

What’s in a Name?

OK, it’s just loose talk - what’s the problem with that? Quite a lot really. Firstly we need to be precise and consistent - meaning matters. The value of a metamodel is it helps separate loose use of language e.g. “HQ” as a unit of organisation from “HQ” - the bricks and mortar. We only have the architectural language allowed by the metamodel. Nothing else is acceptable.

In fact we say system of systems - this suggests that the the overall thing is a system after all. And so it should be.

It is also dangerous because we humans love to split things up, decompose them - reductionism. In separating “systems of systems” from system the next step will be yet another division or specialisation in people, another handbook or process set. What we’ve immediately done is create an organisational interface. Not only is this a weakness but it then encourages us further to think that there is a real difference in how we treat them, the techniques we apply. In one small step we’ve fragmented the very set of people who should be the ones who are charged with synthesis and viewing the whole. [SARCASM]Why not go the whole hog and have ‘sub-system’ engineering and so on? [/SARCASM] Actually we sort of did ...this used to be the case with the format of MIL STD specifications which differed depending on whether the subject was a system or sub-system. It was confusing, caused no end of problems and was utterly pointless.

Being cynical “systems of systems” are seemingly new, shiny, sexy, seemingly impressive to the non systems engineer/thinker and attract money. There is also a lot of political capital invested in emphasising the supposed differences. After all it wouldn’t do to find that we’ve had these around for many years, would it? As Alistair Murray, UK MoD Key Systems Advisor observed in his keynote presentation when talking about NEC, the MoD are very good at jumping on bandwagons. They’re not the only ones. Where funding is available it’s all too easy to do.



Be the first to comment on this post.

Leave a comment

If you log-in you don't need to provide your contact details. Site members can also like/dislike comments and rate posts.

Login | Register




required input

Email is not visible after submission - it is only used to notify you.

My Comment:

Return to Previous Page

All articles/posts © of the respective authors

Site design and architecture © 2010 - 2019 Eclectica Systems Ltd.