Christopher Alexander’s Challenge meets Darren J. Doherty’s Design Process

This post continues and deepens an inquiry into two contrasting understandings of design process.

Christopher Alexander challenges the standard permaculture view of design as a process of assembling pre-existing elements (or parts) into wholes. For Alexander, nature-mimicking or what he calls living design process differentiates pre-existing wholes into parts.

Ironically, Alexander’s approach is a better example of the well-known permaculture principle design from patterns to details than permaculture’s own default design approach, which contradicts it.

While accepting the validity of Alexander’s critique, and the need to develop better patterns-to-details design processes, David Holmgren¹ cautions against construing the two approaches as mutually exclusive. Instead, he (tentatively) suggests whole-to-parts and parts-to-whole modes of design might be construed as complementary but asymmetric aspects of a broader and more holistic understanding of design process including and valuing them both. Asymmetric in the sense that the overall direction is from patterns toward details, but where at times and as appropriate there is also a movement from details toward patterns.

Our last post revealed² that while Dave Jacke’s ecological design process moves primarily from wholes toward parts a la Christopher Alexander, it at times also uses a kind of element assembly (what Jacke calls guild-build). In this sense Jacke’s process embodies precisely the kind of asymmetric relation between the two approaches proposed by David Holmgren.

In this post, we seek further clarity about the distinction and relation between these two design modes, toward our larger goal of increasing the correspondence between our descriptions of design process and the reality of design process. Thanks to Darren J. Doherty, in today’s post we have a rich portion of design process reality to immerse in.

Darren J. Doherty and the Regrarians Platform

Having completed almost two thousand design consultancy projects, Darren J. Doherty is one of the best-known and respected permaculture-associated professional design consultants on the planet. With his Regrarians Platform, he has evolved a design process for generating mainframe farm layouts equally conducive to ecological regeneration and financially viable production systems. You can watch this video sharing where he is coming from if you’d like to learn more about his approach.³

Like Dave Jacke, Darren has evolved (and continues to evolve) his design process at two complementary levels:⁴

• the Regrarians Platform as a general process which might be applied to any farm anywhere⁵
• specific instances of applying the platform to a particular farm

Both levels are essential. On the one hand, the general process only has value to the extent it delivers the goods on the ground (when applied to a particular farm). On the other, without communicating the general process, anyone hoping to replicate it can only copy the shape it took on a particular farm (which might not work elsewhere).

Here we will examine one case of applying the platform both as a window into the process in general and as a bit of design process reality against which we can test our target distinction (differentiating wholes into parts vs assembling parts into wholes) and assess how useful it is or isn’t at making sense of what is going on.

In 2013 Darren completed a whole-farm design process for Yandoit Farm, a 140-acre farm in Victoria, Australia. After a some site visits and consultations with clients Michael and Lisa, Darren sketched up a draft concept design:

What follows is a sequence of sketches and comments representing key themes in the way Darren applied his design process to generate this design. Keep in mind that the exact order is not as important as the overall flow, and that as you’ll see in the clip, Darren moved up and down in resolution or scale and from theme to theme much more fluidly that the following diagrams can hope to portray.

Following on from our introductory comments, we will tune into the reality of what happened inside this process to assess a) the extent to which the two approaches (differentiation and assembly) are present and b) if so, the nature of their relationship.

Entering the Process

Early-on a fellow named Colin showed up with an expensive piece of polystyrene which he proceeded to shake vigorously then throw into the air…

…resulting in this basemap – with 50cm contour lines overlaid on a high resolution aerial photo contour lines (north is up). You can see the existing driveway leading from the top left to the homestead area about halfway down on the left. You can also see a creek running from bottom to top defining the right-hand or eastern boundary.

Darren first sketched in a perimeter fence line something like this (where the west, north and east boundaries were already fenced, and east boundary wasn’t):

If you watch the clip you’ll see that Darren tweaks the details as he goes. Take for instance what happens at 5m:20s, where he fluidly refines the details of the just-drawn main perimeter fence based on the location of a pre-existing (heritage-listed) water channel (or “race”):

This introduces a subtlety with respect to the two different design approaches we are exploring (the differentiation of wholes vs the assembly of parts) and trying to relate to practice. Here’s what happened:

1. roughly sketch in a perimeter fence
2. based on details of the property that are then noticed, modify the shape of the perimeter fence

Three Interesting Points

As simple as these two steps seem, they raise several interesting points as regards the relation between wholes and parts inside Darren’s design process.

In the first step, two parts (a foreground and a background) were distinguished: i.e., the perimeter fence was drawn in. The foreground part then became the focal whole for the next step (as the whole property being designed). Hang on a second – what was a moment ago a part is now suddenly a whole!

So the first interesting point is that the terms “part” and “whole,” as used here to try and describe aspects of Darren’s process, are relative. On reflection, this is true in general:⁶ all wholes are simultaneously parts, and all parts are simultaneously wholes.⁷ What is a part and what is a whole is relative to our frame of reference. So “wholes containing parts,” “larger wholes containing smaller wholes,” “bigger parts containing smaller parts,” and so on, are all different ways of saying the same thing. We are dealing with a nested patterning of part-wholes within part-wholes within part-wholes, all the way up, and all the way down. And it turns out that when we talk of moving from the whole toward parts or from parts toward the whole we are simply discussing two different directions of travel through the same nested hierarchy of part-wholes.⁸ None of this will be news to permaculturalists, given that permaculture is a variety of systems thinking used to working at multiple levels of resolution. But at this stage it is important to eliminate any possible confusion about how we are using the basic terms part and whole.

A second interesting point is that Darren is initially moving from a whole (the wider landscape) to a part (the property under consideration). As just discussed, this part then becomes the primary whole for the rest of the process. Straight away, however, Darren was moving from a part (within this new whole) that caught his attention (a pre-existing water channel) back up to the whole (defined by the perimeter fence), and tweaking that whole so as to better harmonise with that part. Here we start to appreciate the artificiality of any rigid separation of moving from wholes to parts and parts to wholes inside healthy design process. Both are ever present in fluid interplay. As David Holmgren has observed (see introduction), if we try and reduce design to one or the other, we are distorting reality. Our goal instead should be trying to better understand their relationship. In this particular case, tuning into a part (the water race) resulted in some tweaking of the whole.

Another interesting point this leads to is the blatantly obvious fact that the site as a whole already has parts (including the water channel). In other words, the design process does not start with a clean slate or ’empty container.’ It starts with a whole area comprising a current configuration of parts. Now obviously it wouldn’t be much of a design process if these parts were left exactly as they were and nothing new was added. At the same time, it wouldn’t be much of a permaculture design process if the nature and configuration of any pre-existing parts wasn’t appreciated and taken into consideration (whether or not those parts end up being retained, modified, or removed). So whatever else is true of the process, it must simultaneously honour the whole and its existing configuration of parts as well as modifying this configuration and almost certainly introducing new parts to it.⁹

Keeping these three interesting (presumably uncontroversial) points in mind – let us move on.

Delineating the Break-of-Slope

Moving on with his process¹⁰ Darren draws in what he calls a break-of-slope delineation based on the site topography, the location of the existing entranceway to the farm, and the location of the farm homestead. This delineation defines the approximate position of a new drain and driveway weaving through the middle of the property:

To help visualise this break-of-slope delineation, which here takes the form of what Darren calls an in sloped gradient catchment road, here is a profile sketch:

Now in conjunction with drawing in this drain/driveway (herein abbreviated to just driveway), Darren provisionally identifies two dam sites.

Even though Darren is just getting started, this is more than enough new content to hit pause for a moment and consider these last few steps in terms of the assembly vs differentiation approaches.

Different Descriptions of the Same Thing

In the conventional language of permaculture design, we might describe what just happened as follows:

with close attention to the contours, Darren next added the driveway element, re-patterning the connection between the property entry and the homestead. He then linked the driveway and drain so they functioned to passively harvest and direct water to two new dams. Darren masterfully assembled all these elements into a functional whole pattern, stressing its draft status, where the details would be finalised in the act of marking out on site.¹¹

At first glance this statement appears uncontroversial, using familiar language to accurately describe what we have just seen happening.

However, our focus here is the critical re-evaluation of our received ways of understanding and describing what we do when designing. In particular, in light of Christopher Alexander’s challenge, we want to try out his suggested alternative way of understanding what is happening. See if it has anything worth incorporating or taking on board before we dismiss it and get on with business as usual.

Let us first refresh ourselves with Alexander’s slightly unusual alternative way of speaking about these matters:

Within this process, every individual act of building [or in this case farm design & implementation] is a process in which space gets differentiated. It is not a process of addition, in which pre-formed parts are combined to create a whole: but a process of unfolding, like the evolution of an embryo, in which the whole precedes its parts, and actually gives birth to them, by splitting.

…

In the process of differentiation, the whole gives birth to its parts: the parts appear as folds in a cloth of three dimensional space which is gradually crinkled. The form of the whole, and the parts, come into being simultaneously. (1979, p. 365 & 370)

or

Merely additive processes (like the assembly of an erector set from fixed components that are arranged and rearranged) never lead to complex adaptation, or to profound complex structure.

The key to complex adaptation in a generated structure lies in the concept of differentiation. This is a process of dividing and differentiating a whole to get the parts, rather than adding parts together to get a whole” (2002, p. 197)

Acknowledging the unfamiliarity of this language (in particular the word “differentiation” has to be close to non-existent in the permaculture literature),¹² is there any truth or value in this alternative way of describing what is happening?

Is there a sense in which, as he draws in the new driveway, Darren is differentiating space, in the sense of changing some of its shape, and in doing so dividing or distinguishing the whole property into parts?

In a straight-forward sense the act of drawing in this driveway certainly results in two, or more accurately three new parts: the part to the left, or above, the part to the right, or below, and the part in between, as in the actual driveway.¹³

Further, it seems we cannot meaningfully deny that in laying out the driveway Darren created what Alexander calls a “crinkle” in the space – a modification or fold in the three-dimensional cloth of the whole landscape.¹⁴

It is also true, as Alexander proposes is true in general, that the configuration of the driveway and two new dams arise in response to the pre-existing reality of the site as a whole. In other words, the whole of the site came before and defined the context in which the driveway and dams arose. In this sense again Darren is literally dividing or differentiating the whole, in the sense of creating new parts within and from it. The details and configuration of the parts are arising in response to the details and configuration of the whole farm. Though it is certainly changing, the whole is not arising in the sense of coming into existence only after parts have been assembled.

A Dilemma

If we are honest with ourselves, we have a conundrum. Indeed, we find ourselves impaled on the two horns of a rather perplexing dilemma.

On the one hand, we can use standard permaculture talk in its element-assembly sense to meaningfully describe what is going on. On the other hand, Alexander appears to have a point, in the sense that despite the unfamiliarity of his language, everything he says in the above statements appears to be equally true of what is going on.

This conundrum resists the resolution discussed in the introduction – the idea that moving from whole to parts via differentiation and moving from parts to a whole via assembly are two different but complementary actions within sound permaculture design process. For in this case we are not talking about two actions. In the case of drawing in the driveway, for instance, we have just seen that we can describe this same and single action in these two seemingly opposite ways.

We can say that at this moment in his design process, as he drew in the driveway, Darren was adding a part. It is true! He literally added a new driveway! Yes, he wrapped it into the reality of the site as he went. But the point stands – it wasn’t there before, and now it is there.¹⁵ Then, after he added the dams, he had assembled the essential parts or ingredients of a mainframe water harvesting whole system – with a catchment road surface connecting to a receiving and transporting drain connecting to receiving and storing dams. In this sense he has assembled a whole. Who in their right mind could deny this almost embarrassingly obvious description of what happened?

Yet it appears equally true to say, that in that very same act of drawing in the driveway, Darren was simultaneously changing the shape of the existing landscape and making it different, or differentiating it. Further, based on the outcome of this differentiation, he made two further differentiations in the form of the two dams. As depressions in the landscape, these are literally modifications of (or crinkles in) the shape of the land (currently in the mind but when implemented in the ground itself). And both the driveway and the dams came into being and in a sense were thus birthed from within the fabric of the whole farm, which preexisted and then inherited them.

Hence the dilemma. Rather than talking about the relationship between adding or assembling parts and differentiating a whole space as two different types of activities within a design process, we have just discovered that they can be apparently equally meaningful ways of describing the very same activity!

So where on earth do we go from here? How to make sense of this unexpected and confusing turn of events?

Is Alexander’s way of talking about this just a fancy way of saying what all permaculturalists already know? Is there really any need to change standard and pleasantly familiar ways of talking about permaculture design as element assembly? Do we accept one view and discard the other, maybe split into two opposing and quibbling camps? Do we think of them as like the equally valid ways you can describe light as a wave or particle? Do we try and create a middle-ground way or compromise that captures the best of both?

Conclusion, Another Diagram, and Until Next Time

While we take a break and muse on a strategy for steering this inquiry to some kind of meaningful conclusion (in Part Two), we leave you with a final diagram. Skipping much of the process that got him there, this diagram shows a later iteration¹⁶ of the whole-farm layout Darren ultimately generated.

The diagram highlights an overall patterning of areas to remain grassed paddocks and areas to be put into different kinds of forestry systems (you can also see the perimeter fencing, driveway, drain, dam, homestead envelope, and other considerations in the background):

Here we have a complex organic whole in which a generic overall procedure or process has been applied to generate a unique layout relevant to the realities of these people and this landscape.

In our next post (Part Two of this two-part series) we’ll continue exploring the extent to which Christopher Alexander’s challenge might (or might not) help us more adequately describe the details of processes like this, and of sound permaculture design process in general. In particular, we’ll hope to shed more light on at least one resolution of the dilemma we have arrived at.

Meantime, we remain open to your feedback, guidance and advice as to next steps, as well as any mistakes or wrong turns we have made here that we can address next time. Thanks for your interest and support, and see you in Part Two.

Appendix: a Video and a few Photos of Darren’s Design Being Implemented

A clip showing some during and after shots of the main earthworks phase:

Darren’s working draft design sketch:

Darren measuring up the planned driveway with a bendy ruler as part of costing the earthworks:

Michael, Darren and earthmover Graeme marking out with laser level:

Tweaking the main dam details in situ (they’d be under water if they were standing there today!):

References

I thank Darren J. Doherty and James Andrews for their feedback on an initial draft of this post.

Endnotes

1. co-originator of permaculture and originator of the design-from-patterns-to-details principle
2. Though we didn’t fully realise or emphasise this at the time of its publication
3. Keeping mind this video is now slightly dated
4. Dave Jacke has described the two levels as follows: “Certain principles and ‘archetypal’ activities undergird every effective design process, yet each trip through it is unique” – from Edible Forest Gardens Volume Two, p. 142. Christopher Alexander once put it this way: “First of all there are the processes needed to do it [in his case create beautiful buildings]; and secondly, there is the actual doing of it” – From Christopher Alexander: The Search for a New Paradigm in Architecture by Stephen Grabow – p. 22
5. As being laid out in Darren’s “The Regrarians Handbook” (in-progress)
6. as many, many others have previously noted – for instance Christopher Alexander’s “Every part you look at is not only whole in itself, but is part of a larger whole, has wholes around it, and is itself entirely made up of wholes” (The Timeless Way of Building, 1979, p. 483)
7. Arthur Koestler coined the term holon to describe this
8. What Koestler called a holarchy
9. After writing this we happened across Ross Mars making this same point in his book The Basics of Permaculture Design: “In designing, we repackage or re-assemble components already existing on the property and incorporate new ones” (p. 33)
10. Having drawn the perimeter fence and a couple of internal subdivisional fences
11. Keep in mind also that Darren himself at 11m:27s refers to the permaculture principles of “each element performs multiple functions” and that “each element is supported by many other elements.”
12. Upon googling differentiation permaculture design we found that apart from links to some of our own previous posts and a separate post describing some of Alexander’s ideas to a permaculture audience, that the only links were to posts differentiating permaculture from other approaches. The outcome is quite different for assemble permaculture design
13. If we consider that the driveway includes a drain then we actually have four new parts
14. Keep in mind also that Darren himself refers to his sketching in of the driveway as a “break-of-slope delineation“
15. It is also true to say he modified the location of the old driveway, but of course it still stands that he did this by adding a new one
16. Which benefited from ground surveys, the actual construction of the new driveway system, and additional input from other designers including David Holmgren