Is creativity predictable?

The previous section concluded with the introduction of two cognitive terms often heard in relation to design – creativity and invention. These are terms used widely in our everyday lives, and as such are

charged with many misconceptions. There are two schools of thought when it comes to explaining the aspect of creativity. Watson and Perera (1997) but also Steinberg (1994) point out that creativity in design can be much more a social issue when the ultimate judge of the creativity of a product is the society that is using it. Another community considers creativity to be a result of a certain processes and problem solving techniques (Altshuller 1984; Thompson and Lordan 1999; Liu 2000).

There are many questions left open by both communities. Can a creative design process be repeated and still produce creative results? If a creative process is re-usable, are its modifications also considered creative? The counter-argument to these questions claims that it is the deployment of a product in a particular domain that may be judged innovative or creative. The particular solution may be known in other domains, but it is its unusual deployment that gives it the adjective ‘creative’.

Thus, creativity may be associated with both the product of any design method deployed in certain environment, as well as the specific design procedure. Both views have many supporters and opponents. This viewpoint, seeing the product and its deployment as a measure of creativity, has support, for instance, in the community of prototype-based (Gero 1990) or case-based design (Gomez de Silva Garza and Maher 1996; Prabhakar and Goel 1998). These researchers argue that a sufficiently large repository of previous design cases (products) may trigger the awareness of interesting and unexpected analogies or innovative explanations of puzzling features.

Watson and Perera (1997) compiled a list of selected aspects of the case-based paradigm to demonstrate its suitability for creative designs and a close relationship between creativity and the use of previous design cases. They mention several features of various case-based design systems as relevant to creativity. To illustrate their understanding, some of their observations are highlighted in this section. According to the review, a case-based approach to design provides the designers with useful insights as to how components were combined in a previous, familiar case in order to deliver the desired performance of a product. Moreover, reminders of the previous instances of a particular type of design, may improve the current designs by enhancing the creativity of the designers.

However, as Dominowski and Dallob (1995) note, the existence of a familiar representation, whilst solving a problem, may have also an adverse effect on the designer’s creativity. The familiarity may help to clarify the outstanding issues but it may also cause the designer to become fixated on the familiar interpretation. Such a situation typically occurs when a designer tries to re-use the existing approach literally; he tries to conform to the direction set out by the familiar case. Thus, the reference to a familiar design case experienced in the past may inhibit and suppress creative insight, as well as enhance it. From this observation, we may conclude that the presence of the previous design cases is, in general, a helpful but not a sufficient condition for creativity. To this point, Watson and Perera (1997) add that rather than the actual design cases, the process of their re-use is more important. The way, in which the designer deploys the previous cases, influences whether a memory-inspired approach would enhance or inhibit creative element.

There are many models of a case-based approach to designing products. However, there are only a few generic algorithms acceptable as proven methods for designing creative and innovative products.

This may suggest that the view of creativity focusing on the design process (algorithm) that yields a creative outcome is less popular in the design community. However, this is not true. For instance,

Altshuller (1984) developed a theory of inventive problem solving by generalising over a large number of cases that were accepted as innovative patents or inventions worldwide. One of the outcomes of his theory (called TRIZ or TIPS) is an algorithm for improving the chance of coming up with an inventive solution to a given problem.

TRIZ is a very interesting theory because it focuses an inventive design task into a small core of basic procedures. First, an explicit contradiction is identified in the specification of the desired product behaviour. Next, the core of invention is in the removal of such a contradiction. Altshuller understands the contradictions as mutually connected features that are typically observable in a designed artefact, and the improvement of any one of them worsens the other one. So far, TRIZ reminds us of the decisions as described by numerous other researchers (Candy and Edmonds 1996; Cross 1997).

However, the main contribution of TRIZ is that Altshuller generalised many possible contradictions to a matrix of typically contradictory classes of physical properties. Thus, the designer may associate a specific identified contradiction from his problem with one or more patterns in such a matrix.

In addition to assisting with the identification of conflicts, Altshuller also proposed a table of methods and techniques for the removal of the contradictions. He called these methods inventive principles, and most of them were generalised from the previous inventive solutions. Thus, the main principle of TRIZ is that it assigns one or more applicable fixes (i.e. inventive principles) to the feasible and typical contradictions between the physical quantities and behaviours. Nonetheless, it must be noted that the theory has been developed for a particular domain of designing technological products.

Thus, when talking about ‘typical contradictions’ TRIZ refers to contradictions that are typically observable in the domain of engineering design – e.g. if a particular component should exhibit properties of both conductors and insulators or be both present and absent.

Formally, Altshuller’s theory follows the same steps as described by Candy and Edmonds (1996) or Cross (1997). It starts with the identification of the problem (physical contradiction). After finding an explicit contradiction, a set of applicable inventive principles is retrieved from the inventive matrix.

The whole procedure ends with an application of the selected inventive principle and its adaptation for the current problem. TRIZ supports both the retrieval and the application of the inventive principle.

Indirectly, by providing usual sources of contradictions, it also supports the recognition.

The ‘operation’ of retrieving applicable inventive principles is supported by strong empirical evidence gathered from thousands of inventive cases submitted throughout the world. The evidence is presented in a succinct form as a matrix or a two-dimensional table. In the terminology of case-based design, we may say that the inventive matrix provides a referential index for the recognition of certain patterns in the problems tackled. For example, if there is a conflict between improving the temperature and precision of manufacturing, TRIZ suggests ‘creating or stabilising a required structure’ (see below for an example).

The latter ‘operation’ featuring adaptation of the inventive principle for the current problem is supported by examples. A designer may be shown how other applications and implementations of the same inventive principle were designed, what structures or relations they contained, or what they looked like. However, there is very limited help provided with the adaptation or transformation of

rather abstract inventive principles for the particular problem or domain. Consider the following scenario describing a solution of a ‘typical’ inventive problem from (Altshuller 1984):

Problem: “A lens is to be polished to a high precision and accuracy. For this purpose, a coolant needs to be delivered between the polished surface and polisher (made of resin). An attempt was made to incorporate slits into the polishing assembly where water could be squirted occasionally. However, the perforated surface of a polisher performed poorly in respect to the precision…”

Contradiction (verbally): “The cooling capability of a perforated polisher conflicts with its ability to polish glass. A ‘whole’ polisher cannot cool sufficiently, a perforated one does not deliver good results.

Thus, a polisher must be both ‘perforated’ and ‘solid’ at the same time…”

Contradiction pattern: “This is a conflict between a structure exhibiting desired behaviour in respect to temperature and the accuracy of manufacturing.”

Inventive principle: “Create a structure with a desired property (temperature) from the existing one.”

Examples of the principle: “Use interference waves, standing waves, magnetic waves. Alternatively, apply phase transitions, mechanical or acoustic oscillations. Finally, consider cavitation.”

Inventive solution 1: “Make the polisher from a micro-porous material; thus it contains microscopic holes for pouring a coolant and at the same time is ‘solid’ on a macroscopic level. The ‘conflict’ is no more – a porous polisher both is and is not ‘solid’…” Æ implements cavitation

Inventive solution 2: “Make the polisher from abrasive particles frozen into ice. As the lens is polished, the ice starts melting, thus delivering water (a coolant) onto the surface of lens. The polisher is ‘solid’

but simultaneously, water can pass ‘through’ it…” Æ implements phase transition (ice ⇔ water) Nevertheless, as Altshuller emphasises in his book, it is up to the designer to find a suitable implementation of a particular inventive principle for his problem. Thus, the process of recognising and

‘exaggerating’ the contradiction is clearly structured and supported empirically. The next step is left entirely with the designer; he or she must recognise what can be re-used from the different examples and how it can be adapted and implemented. However, the search for the ‘right’ implementation is not blind and uninformed anymore. Already, there is a pattern (model) that needs to be instantiated.

Thus, an exaggeration of a problem specification into a contradictory pair of parameters is instrumental in avoiding the accidental exploration of all possible analogies between the previous cases and the current problem. Instead, a consultation to a ready-made, empirically supported matrix offers a conceptual solution or a model that removes the contradiction and may bring in novelty as well.

As we mentioned above, the TRIZ community (Altshuller 1984; Sushkov, Mars et al. 1995) developed a few techniques for recognising conflicts, and TRIZ itself helps to construct a conceptual solution. However, we believe that the process of recognising such contradictions is not elaborated sufficiently by this community. The recognition process may involve several iterative steps refining the designer’s perception of the problem, and addressing the issue from different perspectives. Another gap that is worthwhile addressing features the fact that unlike in the example above, the contradictions may be sometimes very hard to articulate in such clear and explicit terms. Thus, a designer or inventor may

spend a significant time trying to elicit the essence of a tacitly perceived conflict in various explicit categories before arriving at the ‘right’ one that yields the ‘right’ (i.e. acceptable) result.

To conclude this section, TRIZ (and related research; see also section 5.1.4) has many interesting features that should be contained in a broader framework of design problem solving. As we show later in the document, Altshuller’s theory can be incorporated in the proposed recursive theory of design framing. It may be perceived as one of several mechanisms that implement reflection on design actions.

Nonetheless, we re-iterate that TRIZ was developed for technological systems and explicit physical contradictions. Thus, it is not directly applicable outside engineering design.