Born to Design: Innate Human Behaviors Involved in Learning and Practicing Engineering Design

Abstract

By researching the existing literature for the abilities and conditions necessary for people to successfully solve engineering design problems, this chapter uncovers a consistent pattern of the cognitive processes involved and explains many of the intrinsic behaviors displayed by designers. Limitations in working memory size explain the use of several design-solution achievement devices: pattern matching; early single-solution conjecture; iteration; co-evolution of problem and solution; and intuition. In addition, learning and creating are found to be similar processes, with both requiring and building upon domain experience, in this case actual designing. Similar too are the processes of seeing and imagining, so that von Helmholtz’s dictum that ‘visual sensations are stronger than acts of the intellect’ can be applied to the solving of engineering design problems. This leads to an explanation for another set of intrinsic designer behaviors: a preference for visualizing solutions (over using abstract analysis); single-solution conjectures; object fixation; and found-object designing. Such explanations should help guide future education and research in design.

1. Introduction

In the last two decades, economics and decision-making research has departed the world view of scientific rationalism, both for the inadequate models of actual human behavior it offers and that instructions to follow them do not assure superior outcomes [1, 2]. In both fields, a human-centric position of observation and understanding, instead of prescription, has subsequently been adopted. Research in engineering design appears to be following a similar trend, albeit far more slowly and in a much less unified way [3]. Adding to the initial, modernist aspiration towards a rational design science based on the natural sciences is engineering’s strong association with science, that is ‘applied science’, with the possible inference being that design must be such a science and one in its own right.

As will be shown in this chapter, however, people display common, intrinsic behaviors when designing, and that these both enable and restrict the process of design. It is posited that these behaviors need to be understood so that students can be better guided towards becoming proficient designers.

Towards this goal, this chapter surveys the design-research, decision-making, psychology, cognitive, neurological and child-development literature for commonalities and discusses how four basic human abilities—learning, creating, visualizing and talking—manifest themselves as these intrinsic behaviors in the practice of design.

2. Learning from physical objects important to designing

Engineering design’s ultimate goal is to make or change physical objects, so it should not be a surprise that they are vital to learning and practicing design. In fact, several researchers go so far as to consider making to be the umbrella concept under which design resides (e.g. [4, 5]). Experience with physical components positively affects the path of exploration, discovery and the generation of ideas [6]. Many education researchers acknowledge that only a groundwork of in-hand activities with materials and hardware can prepare one to visualize (e.g. [7], p. 56) and sketch (e.g. [3]) in the engineering domain. Additionally, several recent studies show engineering students’ comprehension of system interconnections to be substantially improved—and abilities to describe reasonable redesign solutions to be increased—after dissecting scaffolded designs, that is performing reverse engineering [3].

The basis of this need to start with concrete examples lies in the principle of connecting existing to new [8]; humans are endowed to easily visualize, but only in terms of what they have witnessed before. Deprived of interacting with parts and prototyping solutions, students will not have any experiences to draw upon when problem-solving in the future—which normally involves visualizing and sketching. No one can interpret, understand or specify functions without having first experienced, witnessed or imagined their implementation in physical form [9, 10, 11]. And, once they do, it is in this form that the function will be automatically visualized [1, 9]. They are not universal absolutes, but derive instead from the actions of physical objects and whose forms the designer will remember. This and other aspects of imagery generation will be explicated further in Section 5.

2.1 Pure abstraction insufficient

Design students with no tactile or visual experience—and therefore without functional knowledge [12]—will use more of their brain regions associated with abstract reasoning when attempting to solve design problems. Having very little knowledge to abstract, however, they cannot progress past the problem-scoping stage to fully develop a solution [13]. Even engineers who are trained analytically only remain so until they acquire experience with the specific problem type [14]; after that they will visualize solutions. Design is thus considered by some as the making of meaning [15].

Flipping through images of engineered artifacts or seeking discourse with a specialist—as substitutes for acquiring genuine, hands-on experience—is not beneficial to design students, for it neither helps them learn nor solve design problems. In the empirical studies of Radcliffe and Lee [16], the authors observed that students scouted design magazines apparently aimlessly, and that their quest to find solutions was restricted by their experience and knowledge. Similarly, Samuel [17] found students were unable to interact with specialists.

2.2 A continual need for making

Since observing, interacting with and making physical artefacts are the sources of visualizing in the first place, it would make sense that these activities are more influential—and producing the associated internal images would require less effort—than using just recall. This conclusion is supported by neuroimaging of design ideation with and without external stimuli [18], and also by Youmans’s [19] study of students in engineering, design and psychology, for how physical prototyping affected fixation on a prior-solution example. In it, product performance improved, and fixation effects reduced, when working (i.e. prototyping) with construction sets was permitted—in comparison with only observing the sets and sketching.

Though expert designers often first envisage then begin sketching, building a model is still critical to successful design [3, 20]. Participants in Crilly’s [20] interviews with professional designers frequently emphasized that the making of models was essential for idea testing. Sketched ideas were rarely considered to satisfy the technically challenging conditions until they were corroborated by physical prototypes. The feedback designers obtained from their own prototyping, Crilly concludes, became in effect an outside critique source, permitting them to disengage from unsatisfactory ideas. In corroboration, Jornet and Roth [15] remark that it is the actual making and the perceiving of the result, which allow for the creation of something new.

2.3 Intrinsic learning structures

However, for this to occur, some sort of pre-existing or intrinsic learning structure must also exist. Daley [21] argues that without it, we would not be able to make sense of our experience, not even at a rudimentary level: ‘Only by having criteria for what constitutes a phenomenon, could we ever proceed to construct any systematic knowledge out of a vast myriad of sensory inputs’ [21]. Ingold ([22], p. 98) corroborates this view, hypothesizing that for physical items to partake in cognitive processes, they must have already been represented: ‘Why should people think with artefacts alone? Why not also with the air?’ So does Helmholtz ([23], p. 32), who states that such a process must be a law of our thought processes which precedes experience, that we cannot receive experience from physical objects without a causation law which already resides in all of us.

3. Creating requires domain experience and imagination

Regarding the neural devices involved in creating something new, Vygotsky [24] notes that the brain’s combinatorial function is not unlike its memory storage function, and is just a further extension of it. He states that ‘The combinatorial action of the brain is ultimately based on the same process by which traces of previous stimuli are stored in the brain, and the only new thing about this function is that, in operating on the traces of these stimuli, the brain combines them in ways that are not encountered in actual experience’. This combinatorial ability is the essence of creativity.

3.1 Definition of creativity

Until fairly recently, however, creativity has had no single, simple definition in psychology [25, 26], with most such researchers now agreeing that it can be considered in terms of the outcomes of novelty and usefulness [27, 28, 29, 30, 31].

The quantity or volume of ideas generated, though not considered by psychology researchers to be a measure of creativity, is embraced as a third requirement for creativeness by some design educators and researchers (e.g. [32]). This definition, which is not entertained here, appears to be driven by a belief that generating more ideas will logically allow the solution of a better idea from the start, thereby increasing the chances of business success. But discerning what is useful requires experience in the domain, and design students cannot be truly creative unless they accumulate practical knowledge [33], just as studies of children’s development show that they are not able to imagine future scenes until they have gained the ability to recall past experiences, typically between the ages of 3 and 5 [34].

Creativity, therefore, is often characterized as an achievement rather than a trait [35, 36]. Yet, at the same time, the creative processes are fully manifest in earliest childhood, and a child’s play is not simply a reproduction of what they have experienced, but a combinatorial or creative reworking of it [24].

3.2 Experience both enables and limits creativity in a domain: fixation

This dependence of creativeness on domain experience has ramifications for designer behavior. Downing [37] writes that ‘The act of remembering is so fundamental to perception and cognition that it seems absurd to argue for a preconceptionless attitude in design. […] the very roots of creative behavior must include the linkage between memory and imagination as a critical part of sustaining ideas, the invention of ideas, or the evolution of ideas in design. The depth and creative potential for memory in design is complex, but should remain a central aspect of any theory of design’.

From the above one can at least surmise, however, that domain experience can limit as well as enable creativity. A quote from Vygotsky [24] is relevant here: ‘[…] if we want to build a relatively strong foundation for a child’s creativity, we must broaden the experiences we provide [them] with’. Since the creative activity of the imagination depends directly on the richness and variety of an individual’s previous experience [24], such an approach should be a continual one for students and experts. This gives direction on how to overcome the so-called ‘fixation’ of designers (on existing artefacts or design solutions) reported by numerous design-researchers and educators (e.g. [20, 38]).

4. Memory devices allowing one to learn and create

Humans use several intrinsic cognitive devices that allow them to process otherwise overwhelming amounts of information so that they can learn, understand problems and generate appropriate solutions. Those devices considered here of relevance to engineering conceptual design are outlined in the subsections to follow.

4.1 A synopsis of working memory

The job of maintaining things in mind while one simultaneously undertakes complex activities like reasoning or comprehension is performed by the cognitive system referred to as working memory (WM) [39]. It is restricted to containing around four ‘things’ at any particular time [40] in both adults and infants [41], each of which is a memory chunk. Chunks in turn are cognitive assemblages, which contain numerous lower-level information elements, and are classified under a solitary function, purpose or schema [42, 43]. They are stored in long-term memory (LTM) and are easily accessed by WM as needed. As knowledge or experience in a topic is accumulated, each associated chunk will represent increasingly more information such that, when learning or practicing in the future, just the single chunk will need processing.

Immense quantities of schemas are stored in LTM and they can be individually brought to WM as necessary. Likewise, information residing in LTM is constantly under revision with fresh experiences and insights [34] per structures which afford exploration and creation [44]. The process substantially changes how individuals tend to approach design problems as they gain experience [14]. Partly because of their information growth associated with learning, chunks can be instrumental in storing and retrieving ‘pre-solution’ models, and recognizing new opportunities and directions for search [45].

4.1.1 Visuals and spoken-language pair in working memory

Baddeley [39] established from experiments that WM has two secondary systems that are associated with the visual and spoken-language domains. One is a ‘visuo-spatial sketch pad’ that orchestrates the setting up and manipulation of visual images. The other secondary system is basically a phonological loop that works with ‘speech-based information’. The two subsystems are the main components of WM in humans, as visual and auditory imagery are their most common [46], with the other imagery types being haptic, olfactory and gustatory.

4.1.2 Verbal coding of images for easy storage and retrieval

The pairing of visuals and spoken language in WM, however, is not likely due only to their frequency of use, as it seems to offer a means of simultaneously accessing—or cross-linking—them. Recalling or generating visual imagery as a consequence of a verbal prompt is commonplace, such as when people are asked if an elephant possesses a short or long tail. Most will answer by recollecting the sight of elephants from memory instead of a list of tail lengths of animals ([46] citing [47]). The process operating here has been extracted from empirical study, wherein Pearson et al. [48] discovered that repeatedly saying a word out loud generated considerably more visualization interference than did secondary visual assignments. This appears to be evidence of people employing verbal coding to expediently save and retrieve visual imagery in LTM ([46] citing [49]).

4.1.3 Working Memory’s information sources: sensory and long-term memory

With easy access to both types, WM information is often comprised of combined sensory and conceptual representations [41], the latter coming of course from LTM. Sensory input to WM is not limited to the visual, of course, and includes auditory and proprioceptive (physical) as well [41].

With regard to its interfaces with perception and with LTM, the WM system operates essentially serially, one process at a time, and not in parallel fashion [50, 51]. These elementary processes—the passing from perception or LTM to WM of information amounting to the equivalent of a chunk—take tens or hundreds of milliseconds [51], and their outputs are held in a WM having a capacity of only a few symbols or chunks. Moreover, the time required for the opposite, to store a symbol in LTM, is considerably longer—as will be discussed in the next section.

4.2 Mitigating WM limitations: general and design-specific intrinsic devices

With WM’s small size, any learning must slowly build upon what are initially extremely simple experiences and issues. At every stage, successive increases in difficulty are bound by these limitations too and consequently can be integrated into the chunks or schemas. Such bounding is evident from Ericsson’s [52] study, which found that students and practitioners should be given cases just immediately above their present ability level. Each such level becomes the next learning stage’s foundation; that is, a person learns only in terms of that which they already know [53]; there should be a real, built-up knowledge framework for the ability to exist.

The time needed to learn, therefore, is substantial. WM’s small size and the time involved in moving items from it to a LTM of effectively infinite size—seconds or even tens of seconds—combine to form the system’s bottleneck [50, 51]. Its effect is accentuated by the necessity of presenting the stimuli multiple times, with definite breaks between ([54], p. 172). Moreover, some critical aspects of the object of the learning must vary while other aspects remain constant [55]([56] citing [57]). The subsequent presentation(s) then allow the brain to creatively consolidate what one is learning ([58, 59], p. 116).

An additional though advantageous corollary of these learning conditions is that object functions are abstracted from their particular forms after seeing many examples [44, 60, 61]. Zeki [61] views this as a crucial step in the effectual acquiring of knowledge; without abstraction the brain would unfortunately be tied to particular objects and specific instances.

Some of the intrinsic means designers naturally employ to deal with or circumvent the limitations of WM size and speed are described in the following subsections. Some not explicitly discussed here include iteration (which will be discussed in Section 5.2), incubation and opportunism.

4.2.1 Matching patterns

In the decision-making domain, which includes design, experts subconsciously and automatically employ pattern matching once they gain problem awareness [62]. They will search for matches between their past experiences and the present situation [62], and, if an exact one is found, execute the normal remedy [1]. If none is found, the experts will then search for similar patterns from their mental collection in the domain to create larger, more applicable ones. This procedure may be assisted with the use of sketches, elaborated on near the end of this chapter. Once it is intuitively deemed suitable, the solution is consciously simulated and evaluated, which involves visualizing how the solution proposal will operate in the situation. The solution is applied if it is judged to work, otherwise it is modified or alternate solutions may be considered. The procedure operates serially, wherein only one solution is conjured and assessed at a time; no comparisons are made between multiple solutions [1]. According to Klein [1], the matching of patterns and the synthesis parts of the problem-solving process are unconscious and intuitive, while the mental simulation part is deliberate and analytical.

4.2.2 Early conjecture of single solution

Ball et al. [63] conclude from empirical studies that the tendency of designers to favor minimal searches for solutions is a reliable one and that basic cognitive limitation—in the face of design’s large information-processing load—is the likely cause. Only a small set of cues from the problem is needed to evoke and conjure a solution, partial or complete, in the experienced designer’s mind [63, 64]. And, as many studies have demonstrated, these designers will usually conjecture then iterate on one solution only (e.g. [14, 65, 66, 67, 68, 69], pp. 145-146).

Empirical studies show that student designers also tend to stop considering alternatives once they have an idea [70, 71]. This tendency persists in spite of the students being recently trained in the methods of rational design [72], where generating and developing multiple design concepts in parallel is considered a best practice (see Section 3.1). Ball et al. [63] conclude from their empirical-study review that novice and expert designers’ strategies for solution-development are very similar, with both groups minimally exploring alternatives, fixating on initial concepts, refining the solution through iteration and patching up concepts for which weaknesses are revealed.

4.2.3 Problem and solution co-evolution

Designers have difficulty clarifying and analysing design problems upfront [14, 73, 74], and the time they spend trying does not correlate with the outcome’s quality ([75], p. 87). In explanation, it is not likely that designers are able to account for all aspects of a problem without proposing a solution [76]. New information and insights concerning a problem may arise from iterative solving, allowing the designer to advance to higher levels of its understanding. As they seek and elucidate information regarding the problem, designers will gain further insight that improves upon its previous understanding. In this way, designers are constantly co-constructing problem understandings and potential solutions [3]; learning is central and inherent to designing [77] and problems and solutions are closely interwoven [78]. No situation has an inherent structure because the problem is learned and understood only in terms of the designer’s own experiences. Figueiredo [79] summarizes co-evolution well, observing that, ‘the process of solving a problem becomes identical with the process of understanding its nature, […] with the information needed to understand the problem depending on the designer’s ideas for solving it’.

4.2.4 Intuition: rapid responses from accumulated experience

Intuition’s primary purposes are expediency and quality of outcome, for it compresses years of experience and learning into mere seconds [80, 81, 82]. It is manifested partly as emotions because accrued experience reveals itself as having a ‘feel’ for choosing a course of action [83, 84]. The process is not, however, an irrational one [82]. And per Dreyfus and Dreyfus [85], such emotional involvement is not found to affect technical rule testing and result in irrational decision-making. To the contrary, Badke-Schaub and Eris [80] propose that basic intuitive behaviors are possessed by all humans, and that they are mainly associated with survival mechanisms.

Intuition is largely based on implicit learning ([62] citing [86]), especially from feedback and the environment. Intuitive processing, though drawing upon experience gained serially and therefore extremely slowly, is parallel in nature and rapidly integrates and organizes complex sets of cues ([62] citing [87])$—$and is right most of the time [1, 62]. Being a subconscious feel for all the factors, their importance and relationships, intuition is in part an extension or furtherance of pattern matching [1, 62, 81].

Conscious deliberation, in contrast, is a ‘low capacity’ channel and can be quickly overwhelmed by large amounts of information. It will only be more effective than intuition when the problem is simple or if there is a clear condition for success [62].

4.3 Learning and creating using same memory transformations

Rather than just supplying memory chunks with additional information, the learning process provides a structured memory, which is conducive to future exploration and creation [14, 44]. The patterns become the basis for future learning and creation, achieved through transformations [8] or, using another term, synthesis [1], the flexible recombination of bits of actual experience to model a new or even hypothetical scenario [88], and the forming of new patterns. From this, it would appear that learning and creating are very similar, which makes sense considering that both can only be performed in terms of that which is already known. Somewhat corroboratively, Kolb ([89], p. 36) states that ‘learning is the process of creating knowledge’.

Dewey ([90], pp. 217–218) wrote that learning can indeed be viewed as a form of creation. He stated that all conscious experience requires some imagination, and that ‘while the roots of every experience are found in the interaction of a live creature with its environment, that experience becomes conscious, a matter of perception, only when meanings enter it that are derived from prior experiences. Imagination is the only gateway through which these meanings can find their way into a present interaction; or […] the conscious adjustment of the new and the old is imagination’. Commensurate with this observation, a refined definition of learning, as guided creating, will logically emerge in Section 5.7.

5. Visualization

In Section 2, it was noted that no one can comprehend object functions without having first experienced their implementation in physical form [9, 10, 11] and that designers first visualize representative embodiment solutions [1, 9] when posed a design problem. That conceptual knowledge in design is always derived from devices and systems rather than abstract concepts—which is the case in science [10]—is due to the brain’s visual areas being exceptionally evolved and capable of inferring meaning to and from real or imagined imagery. The human visual system and its multiple subsystems begin to automatically recognize and interpret situations and artefacts ([61, 91, 92], p. 7), and their internal visualization cannot be avoided once one has become exposed to them [9]. People therefore visualize design solutions in terms of the real-world artefacts they have been exposed to and comprehend [12, 93].

5.1 Shapes have meaning

Through the interaction of the human visual system’s network of subsystems and the learning and creating mechanisms described earlier, and within the limitations of WM, exposures to physical objects progressively allow the inference of meaning and function to increasingly complicated situations and complex artefacts [94, 95]. This is the development of an ability to discover and learn—through transformations—which is tied to and allows the creation of new shapes for problem-solving, so perception is strongly linked to conception [94].

5.2 Drawing and iteration

Since real objects are the strongest influences on us in design, it makes sense to present them—or representations of them such as drawings or sketches—to our powerful, meaning-making visual senses so that we can comprehend and modify them—within the limitations of WM—as we design. This is the essence of the process Oxman [96] refers to as re-representation. A thought which becomes concrete in a sketch or model permits the designer to discover new insights and new ideas [70, 97], and allows emergent features of the solution concept to be recognized, helping the designer make transformative changes [98, 99]. Our ability to perceive includes the distinguishing of structural relationships in design representations [100]. Creativity in design thus may be seen as an ability to transform the externalized representation and re-represent the schema [96]. The overall process is then one of iterations—of making/drawing then observing, and then making/drawing again—and so on. Sketching and drawing are interaction, and not just the emptying of one’s thoughts onto paper through the hand. A restructuring and reorganizing of knowledge result, which, in effect, creates a new prototype (sketch) and a new designer.

5.3 Image generation from both seeing and imagining

The process of seeing is the perception, analysis and comprehension of that which is before the individual—a working assembly of mechanical components, for example—and the associated image generation which occurs in the brain [94]. Foundational shapes, or primitives, and their relationships are discerned, which leads to meaning being assigned to the assembly in memory via transformations and results in generation of the internal image.

Imagining is another process involving image generation, but is solely a consequence of conception, the receiving of information—perhaps transformed—from LTM ([94] citing [101]). Imagining, therefore, derives from either recall or creativity, or both.

5.4 Near-equivalence of seeing and imagining

The above descriptions imply an equivalence—or nearly so—between the processes of seeing and imagining; the image arising in a person’s mind while looking at an object before them is formed by means similar to an image received from LTM and/or WM. It is a view corroborated by many empirical studies [102] and the finding that many of the brain areas that are activated when we recognize and identify objects are also activated while imagining ([102] citing [101]) [103]. Both processes must draw upon memory and transformations for comprehension and image formation. Furthermore, words invoke imagery per the coding scheme described in Section 4.1.2, so a presented object produces an image in the brain in the same fashion that a verbal or written problem in design will. In the former case, the object’s image will be as it is understood in terms of iterated, re-concatenated primitives, in the latter, it will represent the person’s first plausible guess at a solution. The differences between the images will be that the imagined one is not as ‘strong’ as the one arising from seeing and that the imagined image will be constrained by WM’s limited capacity, absent any external means of re-representation and iteration—for example sketching. Such equivalence is further validated by combining two likely possibilities: that all varieties of consciousness use similar mechanisms [104] and; that to be cognizant of some visual aspect of an object, a collection of similar neurons must already be located in the individual’s visual cortex, and their triggering will match that detail in the observed scene [53].

5.4.1 Imaginings form memories too

For the supposition of equivalence to be useful, however, imaginary images should form memories in LTM too, and the plausibility of this is furnished by Downing [37], who writes that an internal image strong enough to make a ‘sensory’ impression in one’s consciousness must be a perception of its own. This would imply that memories made via imagination while designing are as natural as those made via observing external objects.

5.5 Combining past and present using creative imagination

In what we witness and perceive, there is little which is new, and it is evident that we combine past experiences and current sensations to reduce processing costs. If we are in fairly familiar surroundings, there is no need to look closely at the details. The visual circuits stitch old and new pieces together into a contrived, high-resolution experience of the world [105], which imposes itself with overwhelming power and without our being conscious of how much of it is due to present perception and how much to memory [23]. The latter could likely be the larger part [106], with individuals being forced to be highly creative with information inflow because of how ‘impoverished’ their sensory inputs are [105].

5.6 Gathering new experiences and ignoring old ones

The more frequent the same circumstance occurs, and the more its sensation is taken as the object’s normal evidentiary existence, the more arduous an analysis of the sensation will be by additional observation alone ([23], p. 7). That is, the more one is exposed to the object, the less it will actually be ‘seen’ or even noticed, and the effort one has to make to account for it is less as well. This could be a manifestation of the progressive chunking of experiences in memory described previously, and may explain our need to gaze upon new objects for prolonged periods, thereby turning them into old. In everyday life, this permits us to depend automatically on near-instant recognition ([90], p. 52), freeing up processing power to focus on the new. Our sense-perceptions will not regard anything as a sensation if experience-based factors can overcome it ([23], p. 12). An extension of this is the extreme case, wherein past and present are identical, and the experience which results does not enter into perceptual consciousness ([90], p. 218), and nothing is learned.

5.7 Learning a new object is guided creating, so is easier than creating new

The similitude between learning and creating introduced earlier (in Section 4.3) is explored further here, combining phenomenological insights with the somewhat different-leveled predictive perception in cognitive neuroscience [106]. According to Seth [106], everything that is new must be taken in and meshed with the existing to be understood, which is partly a creative act on its own. For example, the observation of art relies on completing the image—that is those elements not seen previously—through the observer’s active participation. They are drawn into the creative act by this involvement, and so experience the joy of ‘making’ which previously had only been the artist’s providence. Seth defines this as a guided process, wherein the observer’s memories and expectations are cast on and into the image. No doubt the same process is in play when generally observing new objects, including mechanical devices; to the perceived object the observer creatively adds components and associated functions as necessary, with the prerequisite being that these constitutive elements have already been experienced and comprehended. Seeing the new—or learning—then, is in some ways analogous to creatively making, and underscores the importance of genuine, built-up experiences in design—even if only observational. The distinction between the two may be evident from Seth’s use of the term ‘guided’, whereby the creative effort needed for learning is inferred to be less than that required for true invention. With its associated guidance, learning is easier than purely creating, since the former just has to match that which is before the observer, in identical terms, whereas the latter has little to follow except abstract constraints and goals. This would imply that creative individuals are intelligent or at least good learners, which empirical research has shown to be the case [107, 108], but it by no means indicates that intelligent people are inventors.

The difference also suggests that the creation of a new object for solving a design problem is more challenging than understanding an existing one, and that, despite learning being a creative act, pure creativity entails more than memory transformations alone.

5.8 Visual sensations stronger than acts of the intellect

Helmholtz ([23], p. 1) observed that no present sensation could be abolished or overcome by an ‘act of the intellect’ (e.g. trying to ‘will’ something out of view) which, together with the above considerations, implies that visual sensations are more powerful than logic or reasoning. If seeing that which is before you and imagining (scenes using memory) are in fact nearly equivalent, the problem-solution imagery generated when working on design problems is then such a sensation, which predicts that instructions to not visualize are likely to be unsuccessful. This likely also partly explains why directives to consider multiple solutions, a standard instruction in the teaching of design (e.g. [109]) have mostly failed, with student and expert designers usually only considering single solutions (Section 4.2.2). Another partial explanation may be that multi-tasking—which such an endeavor will likely involve—is not really a feasible undertaking [110].

Based on the above discussion, a likely better approach to student guidance than procedural directives is the making of verbal suggestions evoking imagery—per the pre-coding discussed previously. Corroboratively, Pugh and Girod [111] write that we often mistakenly only teach concepts instead of engaging students in ideas.

5.8.1 Some well-known designer Behaviors explained

Helmholtz’s observation in Section 5.8 above regarding the strength of visual sensations should apply to external-object fixation as well, wherein the designer cannot refrain from considering an object solution they have recently witnessed. Attempting to think of another solution is an act of the intellect, which of course is weaker—per Helmholtz’s observation described above—than sensations of the actual object. The only remedy, aside from a sufficient time interval, would appear to be the incurring of additional sensations, either one conjured by coded object-wording or by viewing alternate artefacts. It is plausible too that Helmholtz’s observation foretells the phenomenon of found-object designing by students, despite instructors’ best efforts to direct otherwise [3].

In summary, based on the above arguments and the idea that sensation-images can be evoked by coded words (e.g. a design-problem statement), it is possible to account for several designer behaviors through Helmholtz’s visual-sensation-overriding-intellect effect:

• Preference for visualizing solutions (over abstract analysis)

• Conjecturing single solutions

• External-object fixation

• Found- or nearby-object designing

All of these behaviors can only be mitigated, it appears, by adding further visual and tactile experiences to the designer’s knowledge base, or by changing the verbal cues given to the designer.

6. Spoken language

There has been a growing recognition and understanding of spoken language as an important and motive force in the process of design. Encompassing the views and studies of this and other researchers, the following subsections explore how language both affects and effects design.

6.1 Language use in adult designing

Language use in design, with its semantics and grammatical structure, is an active functional instrument and can be considered as ‘doing’ design [112]. It does not serve just to guide another’s eyes across an object; language now takes on an explicit designing function [113]. Design concepts, if they can be described at all, are intrinsically part of spoken language, and so the language starts to structure the process of the design’s fulfillment [112].

6.1.1 Adult designers verbalize design problems to improve solutions

Talking with others, even people who are naïve regarding the design problem, is an effective way of finding a solution ([114] citing [115]). An explanation by Anderson ([54], p. 261) is that a sentence represents a hierarchical group of solution tasks, and so verbalizing a design problem is a way of organizing the design, giving it a hierarchy the way language is structured. Goldin-Meadow’s ([116], p. 24) explanation is similar, concluding that spoken language segments and linearizes meaning that is usually multidimensional [117]. Plausibly, this rearranging can expose voids or create new connections among information chunks, and thus may be an exploratory or transformative means similar to or complementing learning and creativity. Whom we talk with may be important as well, for Goldin-Meadow ([116], p. 139) note that speakers alter their speech according to who their listeners are. Tailoring one’s speech to the comprehension of the audience may force re-examination towards the way they would. Additionally, when talking among peers, domain-specific jargon may represent complicated ideas expediently and at low cost cognitively [118].

6.1.2 Informal discussions most effective

Eckert and Stacey [115] found from their own and other’s empirical studies that the most fruitful personal interactions—in terms of problem exploration and idea generation—are in-person and informal. Such exchanges take place spontaneously without prior scheduling [88], and arise from impromptu groups formed around specific problems [119].

It appears that such flexibility in meetings and communication may allow the product itself to determine the solution structure, just as it relates to language structure. Being allowed a flexible approach when designing—as opposed to following a prescribed, systematic design methodology—has been found to reduce work effort [120].

6.1.3 Storytelling: guidance in design

Stories provide practitioners in several fields with the means of telling what they know without specifying it ([22], p. 110). They do not so much convey pre-set or fixed information but pointers of where to go and what to watch out for. By telling stories, practitioners are able to offer novices general guidance instead of fixed specification; stories are effective at any level in education, which has been corroborated by many studies (e.g. [121], p. 2, [122, 123, 124]). A complete, detailed specification would offer no guidance, and leave the novice perplexed as how to proceed.

In philosophical support, Ingold ([22], p. 110) suggests that the telling of stories has the same roving quality that the practice of making does since they ‘issue from moving bodies and vital materials’.

6.2 Simultaneous language and object-knowledge development in childhood

The results of Lifter and Bloom’s [125] longitudinal study of infants show that their first words’ (FW) emergence is closely tied to the realization that objects are distinct items and have reversible relations between them, for example objects can be pulled apart then placed back together. Irrespective of the difference in infants’ ages when accomplishments in play and speaking were made, there was a consistent relation between them. Lifter and Bloom conclude that these accomplishments were not only a result of maturation, but were also associated with each other via each one’s relation to developments in cognition. The researchers also find that making specific structures corresponds to a dramatic rise in new words, just before their combination. Following this is the advent of first sentences around the age of three. In many studies, a remarkable characteristic of young children’s early words is their referring to objects and events observed in dynamic relationships, such as actions and sounds, and all represent variation and change [126]. This is in fact the ultimate purpose of words according to Dewey ([90], p. 209), being symbols which signify objects and actions. This of course aligns well with the two subordinate WM systems of visuals and language, and the verbal coding of images, discussed in Section 4.1.

6.2.1 Early creativity with words

It has been frequently observed that, from the very beginning of language acquisition, when a young child is unable to find a word which expresses their meaning, they will often invent one [126]. Such constructions, according to Nelson [126], are indicative of the child’s pre-existent conceptual structure—which does not necessarily match that of the adult community. This appears to relate very strongly to the pre-existing structures for learning and creating discussed in previous sections.

6.2.2 Early abstraction

Once the young child acquires a word, it is usually generalized to other ‘similar’ things [126]. This similarity can refer to many dimensions, though only one is the perception of shape and form; others are function and action [126]. This process appears to be the same abstraction designers subconsciously apply to objects and functions with increased exposure to them—as was discussed in Section 4.2.

Prior to this stage, there appears to be no conveyance of abstract concepts; the vocalizations of pre-FW infants are similar to those of chimpanzees, which are considered by many linguists to be no more than expressions of emotional states ([127], p. 50).

7. Conclusions

A survey of the design-research, decision-making, psychology, cognitive, neurological and child-development literature for the abilities and conditions necessary for successful engineering design provides a consistent pattern of the cognitive processes involved. From the survey, four main abilities emerge: learning, creating, visualizing and talking. Learning and creating are seen to be similar processes and using the same memory structures—which explains why creative people are also intelligent—with both requiring and building upon domain experience (in this case, actual design practice) and with both dependent upon and limited by working memory. This limitation explains the use of several design-solution achievement devices: pattern matching; early single-solution achievement devices; iteration; co-evolution of problem and solution and intuition. It is also seen that imagery and spoken language are cross-linked in working memory such that words can conjure images, both real and imagined.

Similar too are the processes of seeing and imagining, such that von Helmholtz’s dictum that ‘visual sensations are stronger than acts of intellect,’ when applied to the solving of engineering design problems, leads to an explanation for another set of intrinsic designer behaviors: a preference for visualizing solutions (over using abstract analysis); single-solution conjectures; object fixation, and; found-object designing. These explanations should help guide future design education and instruction, and also research in design.

Finally, spoken language is shown to not only evoke internal imagery during the conveyance of a design problem, but to assist in the problem’s hierarchical structuring and consequent solution.