What kinds of ‘expert advice’ work in a complex world? Some likes and dislikes

I’ve been talking to a lot of ‘advisers’ in Oxfam, Save the Children and elsewhere recently about what all thisexperts complexity stuff means in practice. Advisers are unsung NGO heroes, repositories of wisdom and experience, working closely with partners and staff on the ground. And those staff typically want to know what they should be doing differently. That’s what experts are for, right?

But if you’re an adviser who has been mugging up on the implications of complexity (see recent posts), this is tricky. How to stop your advice about a systems approach, with its critique of cookie-cutter approaches based on box ticking, best practice, blueprints etc becoming just the latest cookie? ‘Have you done your system/context analysis? Good, now tick the box and move on to stage 2.’

By a very appropriate process of trial and error, here’s where I have got to:

Likes

Asking the right questions: I think you can safely be fairly prescriptive on this. Ask questions about who holds power, who allies with whom, what could be contextual shifts or windows of opportunity that might unblock change. I use a set of guidance questions for analysing change processes that seems to work pretty well. But don’t suggest answers – they have to emerge from the local situation.

Process: Guidance on how to go about designing, evaluating and adapting a change strategy can be useful, e.g. making sure we talk to the people we normally leave out (like faith based organizations, or ex-NGO colleagues with huge accumulated knowledge, who hardly ever get asked in to brief newcomers). But you need to leave doors open to innovation – processes will vary according to context, as will who to talk to.

Case Studies: NGO people prefer stories to formulae, and case studies can liberate the imagination and suggest new approaches.

Signposting, linking and brokering: a lot of the time you are not actually an expert (it’s OK to admit it!), but you know someone who is, or who has been through a similar experience. In practice a lot of advisory work is about linking, rather than directly advising.

Dislikes

Checklists and frameworks: I’m much more dubious about prescribing the content – what a good theory of change looks like, how to strengthen a value chain or empower women etc. OK, this has the advantage of scale and efficiency (tailoring everything to local contexts is very labour intensive), but it is also likely to lead to plans that are either wrong or have big gaps in them

Typologies: I’ve seen the error of my ways on this one. There’s something very exciting about trying to come up with typologies (of types of state, or civil society operating environment, or education systems, or country categories). But as soon as you put them on paper and present them to other people who weren’t involved in drawing them up, creative thought goes out the window as people spend their whole time either trying to decide which box their context fits in, or arguing with the boxes. Hopeless. A compromise is to use typologies in general discussions to expand our general sense of the range of possibilities, but not try and categorize specific pieces of work.

It's perfectly simple.....
It's perfectly simple.....

Diagrams: I may just be graphically challenged, but I am deeply sceptical of those ornate theory of change diagrams that aid types love to produce. The exercise of devising them in the first place can really stimulate the mental juices about how the system functions, what bits are connected and how etc. But then its probably best to shred it, because when someone new comes in and stares in incomprehension at your ‘change diagram’, their brains will go dead.

But Jo Rowlands, a governance adviser colleague and all-round power and change guru, reckons there is something deeper we need to think about – what kind of people do we need to be to do this kind of work?

‘Seems to me that the most useful thing, in the longer term, is whatever it is that gets this new way of thinking embedded into people’s ways of seeing, thinking and doing. Till that happens, people see it as too much, one more thing in addition to all the rest, or the proverbial straw (that breaks the camel’s back), or too difficult, or whatever the block is.

So the important question is how do we help people make that shift so that it becomes second nature? We probably need to be more concerned with the human beings doing both the programme design and delivery, and how they get the chance to become comfortable with it all – so there’s a serious process/people development need attached that NGOs probably need to take on. Plus, of course, masters courses etc.’

So the good news (if you’re an adviser) is that advisory work is still needed, but it may need to be done very differently in complex systems. Over to the advisory community…..

Subscribe to our Newsletter

You can unsubscribe at any time by clicking the link in the footer of our emails. For information about our privacy practices, please see our Privacy Policy.

We use MailChimp as our marketing platform. By subscribing, you acknowledge that your information will be transferred to MailChimp for processing. Learn more about MailChimp's privacy practices here.

Comments

12 Responses to “What kinds of ‘expert advice’ work in a complex world? Some likes and dislikes”
  1. Gerry

    Duncan
    Your first “like” re asking the right questions reminded me of all the coaching training I had as a management consultant (including some really good stuff from David Hemery – yes, the Mexico Olympics 400m hurdles gold medal winner). There are a LOT of materials/courses/trainers/coaches on coaching. I think that the coaching approach possibly starts to address Jo Rowlands’ question re “what kind of people do we need to be to do this kind of work?”. It’s just occurred to me that I need to talk to my adviser colleagues in CARE about this…

  2. Ines

    Duncan, were your reflections only about ‘advising about a system approach’ or more ? are we saying that:
    a. this is the only legitimate ‘advising’ there is? b. that the system approach is the only game in town?

    Either of those seem a bit extreme and to try to replace a dominant paradigm ‘ the cookie-cutter one ‘ with another, the ‘system approach’. It also seems to ignore the fact that advising surely should start where those we work with are at; that it is a mutual relationship where we learn from each other (rather than ‘us’ telling ‘them’ how to think or what to do, where little inputs (whichever direction they go) can open new universes of thinking and practice.

    • Duncan

      Interesting Ines, I guess the point is that to a greater extent than we normally acknowledge, systems and contexts in which we work are complex (in the sense that change is almost impossible to predict, and very hard to attribute to any particular intervention). As you say, in such complex systems, starting by listening to the people we work with (whether staff or partners) and then asking intelligent questions etc is undoubtedly better than transmitting ‘wisdom’ from head office.
      But there are limits to that. We should of course learn from each other, but advisers are called advisers (and paid) for a reason, so surely it would also be misleading to say ‘hey we don’t advise, we are merely here to listen and learn’, no? The question is what kind of advice/guidance works in complex system and what doesn’t.

  3. Sharad

    Guidance approaches should draw from ontology. It would be the source, the base or foundation, for the attitude towards providing guidance. Indeed the coaching arena provides a rich source of knowledge in this respect.

  4. Thalia Kidder

    Thanks Duncan and Jo Rowlands. In a few words: I agree that the ability to ask good questions is often much more valuable than the ability to tell people what you know… even if you can communicate well. Good questions can get others to recognize the ‘boxes’ we’ve been stuck inside, come up with our own answers, and think and act more broadly. [Much better than checklists, framework and diagrams!]

    The issue is that questions can ALSO limit what’s legitimate to ask about, so that’s not a magic solution either. I’m sure each of us has our list of the ‘best questions’… for me, they need to be ones that
    – make power relations explicit
    – make visible what we consider ‘legitimate and normal’, and challenge that.

    • Duncan

      Good point Owen, one question that sometimes helps is ‘how will you know whether you’ve succeeded, and how would you convince a sceptic that that is the case’. That leaves it open on what kind of evidence to collect in a given context.

  5. Alix Tiernan

    Picking up on Jo’s question about what kind of people we need to BE to do this kind of work, I’d like to suggest that complexity thinking needs skills that not all people working in development have, and may not necessarily acquire through guidance and training. People working in development are wonderfully diverse, and have a very broad range of skills that they have been employed to use. A lot of the thinking around complexity that we are doing is based on the assumption that everyone will ultimately be able to use complexity thinking and that thus all development work will ultimately be improved and greater impact will be achieved. I personally often feel frustrated by the fact that the intellectual skills required for implementing this approach are probably only held by a small minority of development workers. We might need to simplify complexity!!

  6. At a basic level, I would suggest that the role of such internal advisors should be to help identify whether in fact an issue or problem falls into the complex/chaotic category or not. From there I suppose it gets a bit more challenging in terms of operationalising complexity thinking within programming processes. Especially for INGOs who are very risk averse (and understandably so given the intense scrutiny they receive).

Leave a Reply

Your e-mail address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.