Testing a Different(iation) Approach to Permaculture Design Process – Part Three: Another Example

Note: We dedicate this post to the memory of renowned permaculture author and teacher Toby Hemenway. Toby was the author of the best-selling permaculture book in the world (Gaia’s Garden) as well as his more recent The Permaculture City. He passed away yesterday. We feel blessed to have had Toby’s participation and supportive comments on Making Permaculture Stronger. Indeed, this post and the two last posts were directly prompted by a comment he made (as you’ll see in our conclusion below). Rest in peace Toby. This post is for you.

In the last post we shared an experimental example of permaculture design conducted as a process of differentiating a pre-existing whole into parts (as opposed to assembling parts into a whole, as design is more typically defined and discussed in the literature of permaculture design).

In this post we share another example.

On the afternoon of June 30, 2016, we talked, walked and worked with a friend, Adam, on the overall design of the five hectare (12.5 acre) block of land Adam shares with his partner Tink.

Though we’d done some paid consultancy for Adam and Tink on this property in the past (focused more around the house), we were on this occasion visiting as friends, and just having a bit of a casual play with ideas. This gave the session a relaxed feel that was most conducive to how smoothly it all unfolded.

The land is located on New Zealand’s Kapiti coast, about an hour’s drive north of Wellington:

Here is a basemap of the property (including contour lines at 50cm intervals – yep we’re definitely on the side of a hill!):

During this session our focus excluded the home and surrounding gardens which you can see here, along with the most prominent neighbour’s home:

Our focus was the grassed paddock areas comprising the bulk of the property:

Tuning into the People

Clarifying a Destination

Sitting in the dining room with a view out over the space, we started by reviewing Adam and Tink’s plans for this part of the property.

We didn’t even write it down, but the spirit of what Adam shared was something like:

A beautiful, manageable mixture of open & treed areas making places animals including ourselves can happily frolic, wander, gather.

Desired Areas

The wishlist was straightforward:

  • Trees for shelter, fuel, food, fodder, birds, beauty
  • Grazing areas (for an anticipated future menagerie of livestock)
  • Possibly a dam as a sensible use of the valley

Tuning into the Site

As per what you’d expect in any permaculture design process, we next tuned into some of the most significant aspects of the site that the design configuration had to factor in. Here are some of them.


As you might expect from viewing this wind rose from Wellington1 which is just down the road and one of the windiest places in the country…

…strong winds from the north were a big deal on this site (note that the shape of the hill the property is on blocks the strong southerlies evident on the wind rose):

The Valley

The valley running across the property toward the north-west as shown here was the most prominent topographical feature.

The Views

The shape of the land combined with the location of the window of the room (dining room and kitchen) Adam and Tink spend most of their inside time created an interesting pattern of parts of the property visible and not visible through that window:

Moving to the neighbour’s views, they enjoyed ocean views across Adam and Tink’s property which Adam and Tink wanted to honour and maintain.


To simplify, the soils were richer on the 3/4 of the property closest to the home, then their quality dropped off down the back.

Unfolding the Design

Getting Started

We started the unfolding process by having a bit of play with a pencil on a print-out of the basemap:

Let’s now unpack this and show an approximation of how the design was unfolded using a step-wise differentiating process. As mentioned above, we at the very start had differentiated the property at large from the homestead area as shown here. Our aim was to zoom into what is here numbered 2:

We then initially differentiated this focal area into three main sub-chunks: the west-sloping area below the home, the valley running north-west through the property, then the remaining extent, which was a very subtle ridge also facing north-west:2

As this threefold pattern emerged, we earmarked Area 1 for proximity-to-house-appreciating market gardens, future dwelling locations, and general amenity landscaping possibilities. We also tentatively, and without much regard for scale or even the contour lines, further differentiated the valley into a dam wall (2a), a retained body of water (2b) and a ribbon of dam-side plantings (2c):

Under our gentle encouragement, Area 3 in the above diagram then split itself into an area for grazing (the area closer to and visible from the house) and an area for a woodlot (respectfully numbered 3 and 4 below).

Almost in the process of differentiating out the woodlot area, this area further subdivided itself into three bands – 4a for relatively shorter trees, 4b for medium-height trees, and 4c for the tallest trees:

This differentiation was based on the idea of making that entire swath of property visible from the house (as defined by the taller, medium and shorter trees which in this configuration would all be visible) as well as increasing solar access for all three tree heights due to the approximate orientation of the bands with respect to north.

At about this stage we went for a stroll with the intention of finding and doing something about the flaws that our desk-top doodlings were sure to contain aplenty.

The first thing we discovered on a closer look is that a dam was not a viable option for the valley. Not enough flow to justify the expense and effort of building it and not an issue to get a vehicle across during most months of the year. So we nixed that idea and decided to instead look at plantings to accentuate this relatively lush and sheltered area (more on which later):

A second discovery was realising that the extent of the area we’d earmarked for the woodlot felt excessive and that we could shrink it down to allow a more open area (numbered 4 in the above diagram) that might be open pasture or comprise widely-spaced deciduous trees than in future could be grazed between and underneath.

A third development was realising we’d neglected to harmonise with the preexisting shelter plantings of (young and somewhat struggling) natives along much of the two long sides of the property. So in the below diagram we link the woodlot in with the idea of thickening those up (in the above diagram labelled 5d & 5e).

Further Iterations

About here we moved on from our roughly scribbled jottings onto the computer where as well as getting things more to scale, we kept iteratively transforming the configuration towards being better adapted to place and people.3

Here is as far as we got, all based on gradually improving from the starting steps shared above:

In getting to this point, our guiding question continued to be “what issues does this configuration either leave unresolved or actually create?”

One unresolved issue was how to enhance and work with the valley area. What came up here as an alternative to the unviable dam idea was differentiating the space into three new sub-areas – a moisture loving and beautiful deciduous tree (such as weeping willow) clump in the low point, flanked by two clumps of productive and beautiful large deciduous trees needing fertility and drainage (such as walnuts and chestnuts). These trees would have to exceed 20m in height before they even started to impede the view of the main grazing area from the house (in effect, along with the configuration of the rear woodlot plants, the whole property would eventually be visible from the home, even if some it only as tree tops).

An additional example was that in thickening up the perimeter shelter belt along the south-eastern boundary, we would both be putting effort into shelter where shelter was not required and potentially encroaching on the neighbour’s ocean views. So as you’ll see in the diagram above we left the natives unthickened on the south-eastern boundary but left the thickening in place on the north-western boundary (where the trees could get to at least 20m in height before compromising the neighbour’s view – until then they would actually be blocking the view of an unsightly new highway at the bottom of the hill).

Another example was realising that with the addition of a further strip of trees running along the south-eastern boundary between the valley and the home, we could complete a long sinuous wildlife corridor extending from the so far unmentioned ‘tortured'(poor soils plus extreme winds) forest south-west of the property (as you can see here if you’re interested). This would also provide a fully shaded walking path for summer meanderings.


So ends the second report of an experiment in explicitly conducting permaculture design as a differentiating process. This is not to say that we permaculturalists do not already do something along these lines. Indeed, we almost certainly do. We just don’t tend to say that is what we do! As Toby Hemenway has put it:

I think Alexander’s concept is much closer to how permaculturists actually design, by starting with something that is already a whole and then differentiating and integrating additional factors into it. The issue is mostly that our language has not caught up to our practice… …I’m not surprised that permaculture is taking a few decades to figure out what we do in practice.

Toby’s observation accords with our own impression, as documented previously, that the literature of permaculture design has emphasised the idea of design as an assembling, joining, or connecting process. This emphasis has obscured the fact that the reality of sound permaculture design process, when it is embodying the principle of designing from patterns to details, is more accurately construed as a process of transforming a pre-existing whole by differentiating the one whole into two parts, and then further differentiating those parts, and so on.

We would love to hear from any permaculturalists out there reading this as to how it is all landing for you. Specifically we’d love to hear your answers to these questions:

  • Do the two design process examples we’ve just shared resonate as consistent in general terms with how you would describe the process you use, have read about, or have been taught? If so, how so? If not, how not?
  • Are you aware of any existing design process examples or definitions in the permaculture literature that explicitly discuss design process in this way?4 If so please let us know where they are hiding and let’s acknowledge them!
  • Do you feel that this entire inquiry/discussion is contributing to the goal of making permaculture in its design system sense stronger, better able to generate nature-mimicking systems, internally more consistent? If not, what sort of work would you see as better serving this end?

Thanks for reading, thanks in advance for any comments you’d care to share, and in our next post we’ll review and bring to a close this inquiry circuit, which has been the first of many. We’ll then be returning to Making Permaculture Stronger’s big picture and share and inviting input on plans for where we’ll head next.


  1. Courtesy of www.niwa.co.nz
  2. you can see a dashed line representing an already-existing fence below the number 3 that happens to coincide with the ridge line
  3. the above computer-rendered diagrams being generated after the fact to try and make the steps in the process more visible
  4. Dave Jacke is a rare example as we’ve discussed in a past post


  1. Feels intuitive and fluid. Dan do you think the resultant design is somewhat a function of the designers assumptions/ skill/ knowledge (and I suppose on the negative flipside, bias/awareness) and relies on that background intuition, or could it be done by anyone but would just take a bit longer?

    1. Thanks Sharn – and great question. I’m not sure, but no doubt experience brings a lot (which as you note can be a mixed bag). I often wonder if a bunch of different designers fluent in the same process understandings would generate the same designs, more or less, for a given project. That is my hunch, but it would be a fun and worthwhile experiment to run. I’m personally finding these days that rather than being the indispensable design expert, a large part of what I bring is helping the design team to focus on the right thing in the right order. When this happens and at any moment we are focusing on just one decision, I find that pretty much everyone makes it similarly, though that said having someone on hand who can help make sure all the necessary influences on that decision are made visible adds a lot. I’d be interested to hear any thoughts you or anyone else has about this…

  2. Nice post, Dan. Great visual presentation really shows how the process evolved. Toby was right – it will take some time for the language, the practice, and the science to all get on the same page. PS – I was touched by Adam & Tink’s desire to respect and maintain their neighbor’s ocean view – very observant, and a great example for others to follow.

    1. Many thanks Michael that’s awesome to know – though just a start I’d love to see not only more documented examples of permaculture design, but that move toward some kind of common format such that the reality of the process generating the outcome becomes more visible. And yep, I’d totally want Adam and Tink for neighbours! By the way your site http://directoryofpermaculture.com/ is awesome – great work!

Leave a Comment

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