Bisociation - the Fractal Method

Keith Hudson

'I have yet to see any idea, however complicated, which when you looked at it the right way, did not become still more complicated.' Poul Anderson

There are several excellent (and enjoyable) techniques for developing new ideas in groups - for example, synectics or brainstorming - but very few have been described for use by the individual working alone. However, whether concepts are created by individuals or groups, the process is exactly the same and has been described by many writers through the ages (for instance by Aristotle in 'Prior Analytics' and Koestler in 'Act of Creation').

'The central principle is to find common connections between separate objects (or concepts). Koestler called it bisociation'

The central principle is to find common connections between separate objects (or concepts). Koestler called it 'bisociation' and that is the term I tend to use. It is also known as lateral thinking (Edward de Bono) or divergent thinking (Liam Hudson) but these are essentially concepts themselves (and an encouragement to 'free up') rather than techniques in their own right.

However, there is a problem with bisociation. Before making the connections, do you analyse the objects (or concepts) to the finest detail or do you use the objects (or concepts) in their entirety? If you do the first (as the cortex automatically does in the process of perception), then you end up with such an astronomical list of elements that the necessary bisociations take impossibly long to perform in a conscious and systematic way. On the other hand, if you try to bisociate complete objects (for example, an elephant and a lawn-mower) or complete concepts (for example, democracy and neo-Darwinism), you have a task that may take you weeks or years. (And, unless you are a philosopher, it is better not to use concepts as starters anyway.)

Some sort of intermediate method of analysing (or fractionating) objects is therefore necessary in order to arrive at a procedure which is both productive but not too tedious or lengthy. The entities that I use are called 'fractals'. These are shrewd guesses from cognitive science as to the partial stages that are involved in the perceptual and conceptual processing in the cortex (for more about fractals, see the note at the end of this article).

In this summary, I will only give a partial list of some of the fractals I use. These are some of the more generalised ones taken from the three main analytical areas found in the cortex:

Visual

(1) Shape (overall, distinctive parts, size, etc);

(2) Colour (overall, distinctive parts, luminosity, shading, etc);

(3) Distance (causal and spatial relationships with others, etc);

(4) Movement (direction, speed, locomotion, etc);

Somatic

(5) Tactility (touch, texture, pressure, temperature, sharpness, etc);

(6) Muscularity (effort, weight, posture, (a)symmetry, etc);

Auditory

(7) Loudness (power, percussiveness, etc);

(8) Pitch (tone, quality, harmonics, etc);

(9) Naturalness (musicality, aesthetics, acceptability, etc);

(10) Communication (speech, code, threat or other emotive signs, etc).

'It is very important to choose a pair of objects at random and not to make prior judgements as to their suitability'

Let us now look at the Fractal Method in practice. First of all, it is very important to choose a pair of objects at random and not to make prior judgements as to their suitability. An excellent method is to choose two items at random from an encyclopaedia or an illustrated dictionary (the Oxford Children's Illustrated is ideal - the colour photos and artwork are superb). The next step is to identify fractals in each object and, taking each fractal in turn, to find the common connection - however nonsensical it may be. Write the ideas down as they arise in 'stream of consciousness' manner with no attempt at evaluation. (That will be done on a separate occasion.)

'I had been cutting the lawn and I determined to bisociate the lawnmower with the first object I would see on television when I went indoors. It happened to be a circus elephant!'

Here is an example from real life. I had been cutting the lawn and I determined to bisociate the lawnmower with the first object I would see on television when I went indoors. It happened to be a circus elephant!

What follows are the actual notes derived from Fractal 1:

'The shape of an elephant - so different from a lawn-mower! - how would you put handles on an elephant? - or ears on a lawn-mower? - the trunk of an elephant is its most distinctive part - what about that of the lawn-mower? - the grass-collecting box, I guess (if only because mine is falling apart) - all right then, let me bisociate a trunk with a grass-box - what does a trunk do? - well, in circuses, they'll suck up water and then spray it over the clowns - or they can wash themselves - they can spray each other - water shoots out at great speed, long distance - could this be done with grass chippings? - why not? - suck them in a trunk-like tube and spray them away from the lawn...'

And so on - part of several pages of notes that I made in the next hour or so. Typically, two random objects, when analysed into fractals and then re-combined, will produce about 200 partial ideas and 30 to 50 complete ideas in a session of a couple of hours or so.

'Starting with the first random pair ('isthmus' and 'chicken coop'), and choosing others when I needed them, I developed over 4,000 partial ideas'

What are the disadvantages of the Fractal Method? Firstly, it is hard mental work (particularly in the first ten minutes or so, but it gradually becomes more enjoyable). Secondly, when you look at the results on a later occasion, 90 per cent of the ideas will be seen to be impracticable, not to say ridiculous. But do not be surprised about this - this occurs for all creativity methods. (However, there are always compensations. Ideas which seemed laughable when they were originally created often take on a practical air during assessment.) Thirdly, of the remaining ten per cent (the ones that look sound and practicable), you will discover that most of these will have already been discovered. (And, indeed, some of them will have been used for many years! For example, looking through my pile of notes, I saw that I had virtually rediscovered the combine harvester!) Fourthly (once again, this applies to all methods) more ideas for tangible objects and products are produced than for intangible services or concepts.

Despite the fact that the human brain is capable of the most abstract ideas (for example, mathematics or quantum physics), nevertheless, it remains the case that our cortex is happiest when 'objectifying' things. This stood us in good stead during most of our evolution, so it is a strong bias.

However, to summarise, the big advantage is that, by using the Fractal Method, and given sufficient motivation and time, an individual working by himself or herself can systematically create unlimited numbers of ideas. I once carried out a week-long test of this method (two three-hour sessions per day) using my favourite children's illustrated dictionary. Starting with the first random pair ('isthmus' and 'chicken coop'), and choosing others when I needed them, I developed over 4,000 partial ideas, over 350 complete ideas, and consumed half a ream of A4 paper and three ballpoints. After a day's rest and two more days of evaluation, I ended up with five reasonable-looking ideas, of which two were useful in other areas of my work and three were practicable ones. Two of the latter were handed on to an inventor friend and one of these, I gather, now has commercial promise.

So the Fractal Method, while practical, does have its cost - time and energy. As the economists say, there is no such thing as a free lunch.

There is, of course, no reason why the Fractal Method should not be used in group sessions, too, so long as there is a good rapporteur or secretary. I should be delighted to hear from any individual or group who tries this method out - I am on the look-out for successful examples for a book.

Footnote re fractals: The word 'fractal' cannot yet be found in a dictionary because it has a very recent conceptual origin, and, even now, is in the process of clarification. It shares some part of its meaning with 'fraction' - that is, a fractal is a part of a whole - but it also contains the richer notion that it 'enfolds' some significant part of the meaning or form of the whole from which it has been derived. A fractal can be regarded as both a 'reductionist' and a 'holistic' version of the whole. Thus, a branch of a tree can be considered, not as a 'chopped off' fraction of a tree but as a sort of tree itself - containing many of the features and functions of its parent tree.

'Human consciousness may be considered to be a fractal of the universe'

Fractals are beginning to turn up in many areas of science - both practically and conceptually. In computer problems, recursive programming can be used to describe solutions in terms of easier or smaller versions of the same question. In graphic terms, a large curve can be described in smaller curves. An ice molecule is a fractal of a snowflake. In mathematics (differential calculus) an integral is a fractal of an equation. In biology,a DNA molecule is a fractal of the mature life-form that will ultimately develop from it. In a hologram, one small area of it contains the essence of the whole. In modern physics, a single quantum event (such as the release of a photon from an electron) 'enfolds' almost all the fundamental problems of philosophy (such as causality and meaningfulness) so that modern 'hidden variable' theoreticians such as David Bohm are seeking even 'deeper', 'sub-quantum' fractals that also contain the essence of the whole (the 'implicate order'). Human consciousness may be considered to be a fractal of the universe.

Similarly, the fractals that are used in my Fractal Method have the simultaneous meaning that they are reduced parts of the whole (so that enough permutations can be manufactured between them) but also retain sufficient conceptual relationship (with the whole) so that what is produced - by way of ideas - are practical rather than highly abstract (and thus not immediately usable). The Fractal Method can therefore be considered as a pragmatic 'middle way' between the sorts of comprehensive, fundamental ideas that only turn up once every few decades (such as evolution and relativity), and the infinite number of trivial ideas that can be produced by permutating the smallest constituent parts of the original items chosen.

The Fractal Method gives us some freedom in partially unravelling our existing, tightly constrained mental concepts and allowing re-combinations. There is absolutely no reason in principle why this should not be aplied to social innovation as well as to conceptual or product innovation.

Keith Hudson, 6 Upper Camden Place, Bath, BA1 5HX (tel 0225 442377; fax 0225 447727).

Previous / Next / Table of Contents