⚡11. Conclusion
Marcos Antonio de Lima Filho, PhD.
11.1 Contribution to Knowledge
This study began without a preconceived theoretical framework but rather with an open interest in the competitive design of consumer electronics. In this setting, “evolution” quickly emerged as a core category. A cornerstone of the grounded theory methodology is the principle of constant comparison, which in this context led me to contrast the development of smartphones with that of commercial aviation and biological evolution. This potential analogy suggested a categorisation of innovation as disruptive, directional, stabilising, and purifying. These patterns give rise to a punctuated model for the evolution of designs, industries, and technologies.
These findings provide insights into the questions that emerged during the course of this research. One of the main questions was whether the evolution of designs is comparable to the evolution of species. Additionally, the study aimed to determine if there is empirical evidence supporting this analogy and which principles of evolutionary theory can be extended to design evolution. Furthermore, the research sought to understand how these principles could enhance our comprehension of design evolution. By examining multiple generations of smartphones and commercial aircraft models, the study concluded that there are indeed patterns of design selection that transcend specific industries.
In conclusion, the evolution of industries is marked by more than just the disruptive and radical; it encompasses cycles of disruption, diffusion, consolidation and obsolescence. The following subsections present a summary of these research findings and reflect on their implications for design theory and research on disruptive innovation.
Summary of Findings
The discussion of the thesis begins by examining how Clayton Christensen revolutionised the research on disruptive innovation and the possible drawbacks that followed this redefinition (Section 6.2). The emergence of smartphones and the history of aviation pose several challenges to Christensen’s model of disruption. Under this paradigm, major technological advancements such as the iPhone or the Jet Age are no longer considered disruptive but rather recognised as “sustaining innovations”. As a result, Christensen’s disruption theory has disconnected itself from the understanding of previous scholars and the common conception of disruption among practitioners.
The iPhone’s release in 2007 was a major milestone in the mobile phone industry. Despite initial skepticism from incumbents and critics, the iPhone’s architecture, featuring a touchscreen and no physical keyboard, became a reference design for the industry. However, Christensen’s theory of disruption classifies high-end innovations like the iPhone as sustaining innovations rather than disruptive ones. Indeed, Christensen himself predicted that Apple would fail with the iPhone, but it proved to be a phenomenal success, leading Apple to become the world’s first trillion-dollar company (Section 6.3).
The changes observed in this industry can be better explained through an alternative paradigm of disruption. The introduction of touchscreen smartphones in 2007 represented the emergence of a new dominant design in the product. Prior to this, mobile phones primarily relied on physical buttons, compact keyboards and keypads for user input, and the market was populated with a variety of product architectures. The introduction of touchscreen smartphones disrupted this design paradigm, replacing traditional input methods with a large capacitive touchscreen, multi-touch functionality, and a virtual keyboard. This new product architecture has gained a growing allegiance from consumers in the marketplace since its introduction. As a result, adoption has reached above 95% of handset models since 2014. Therefore, despite some considering the iPhone as a sustaining innovation, it can be seen as a disruptive innovation when viewed through the concept of dominant design (Sections 6.4 and 6.5)
Christensen also admits that his model of disruption fails to explain the market changes that occurred during the introduction of jet engines: “These are anomalies that the theory of disruption cannot explain” (Christensen & Raynor, 2003, p. 69). Unlike his peers, Christensen reclassified the jet engine as a sustaining innovation, given that jet engines were neither cheaper nor technologically inferior to piston engines (Section 6.6). Early concepts of disruption, such as those proposed by William Abernathy, provide a different perspective. Abernathy defined “conservative” innovations as those that extend or refine existing design concepts, and “disruptive” innovations as those that destroy established concepts. Several scholars shared this understanding, readily recognising the radical and disruptive status of the Jet Age and the jet engine (Section 6.7).
The rise of Boeing and Airbus, both based on superior technologies, also challenge the premise of low-end market entry of the disruption model. In addition, the decline in regional jet sales further undermines the supposed predictive power of Christensen’s model. To demonstrate the superiority of his theory, based on its ability to generate falsifiable predictions, Christensen predicted that regional jet manufacturers such as Bombardier and Embraer would disrupt the Boeing and Airbus duopoly. However, such disruption did not materialise. Instead of growth, the regional jet segment actually lost momentum and shrunk to just one-third of its size in 2004 (Section 6.8).
The discussion in Chapter 6 ends with a proposal for an operational definition of disruption, which aims to provide a more precise criterion for assessing technological advancements. In sum, I propose that disruptive innovations can be seen as the introduction of new qualitative traits, while incremental innovations involve gradual improvements to existing traits (Section 6.12).
To create disruption, practitioners can take two approaches. The first approach is to introduce new qualities, such as new components or features, into existing product architectures. Alternatively, they can focus their efforts on pioneering entirely new product architectures, dominant designs, or product categories. Choosing the right approach depends on understanding the current stage of industry evolution. In early stages, industries are particularly open for new design concepts, as a dominant design is yet to emerge. However, as a dominant design emerges, the first approach becomes more relevant.
This concept aligns with the etymology of the term and corresponds with how practitioners commonly understand disruption. On the other hand, the definition presented here is at odds with Clayton Christensen’s definition. Unfortunately, Christensen constrained disruption by embedding in it a specific business strategy intended to address a resource allocation problem (The so-called “dilemma”). Grounded in data, this research proposes instead a more generalisable characterisation of disruption. This aims to reintegrate disruptive innovation with its evolutionary research tradition, as pioneered by the works of William Abernathy and James Utterback.
The Chapter 7 discusses the concept of directional selection in nature and technology, highlighting the importance of incremental innovations in product development. It examines two recent studies that tracked directional selection in Pacific salmon and bighorn sheep (Section 7.2), contrasting these instances with the incremental evolution of smartphones and passenger aircraft. Directional selection is akin to incremental innovation, which is commonly understood as gradual enhancements of existing features and functionalities. Term “directional” is used to emphasise the evolutionary analogy.
Incremental innovations, although often overlooked, have a significant cumulative effect on product cost and performance over time. Disruptive innovations, on the other hand, introduce new qualities, features, or product categories. Both types of innovation are necessary for maintaining economic viability and a competitive edge in the market (Sections 7.3 to 7.5). The transition from disruptive to directional innovation is a natural part of industry and product evolution, with incremental innovations capturing and enhancing the potential created by disruptive innovations (Norman & Verganti, 2014).
Incremental innovations have been favoured in aviation due to the cautious nature of this industry and the financial risks associated with introducing major new aircraft. Manufacturers and industry stakeholders prefer incremental upgrades that build on existing aircraft in production, as major new aircraft programs can be costly and risky (Section 7.7). Advancements in engine efficiency have been crucial in reducing fuel consumption, contributing to the popularisation of air travel by reducing operating costs and ticket prices. This research reports that modern aircraft are 72% more efficient than the aircraft produced in 1960, thereby contributing to a 40-fold increase in passenger traffic. Multiple studies confirm this trend, although with some variability in reported fuel burn reductions (Section 7.6).
Similarly, incremental innovations have played a crucial role in making smartphones more affordable and accessible to a wider user base. There have been directional changes in various measures, such as display size and resolution, computational and graphic performance, battery capacity, memory, storage, and camera sensors (Section 7.8).
Chapter 8 highlights that stabilisation is crucial for the long-term survival of both living organisms and designed objects. Just as in biological evolution, there is evidence that most traits in technological evolution remain unchanged, while only a few undergo directional or disruptive selection (Section 8.1). However, the widespread emphasis on disruptive innovation has led to a pro-disruption bias in academia and the business world. This bias tends to prioritise disruptive and radical innovation concepts while disregarding other forms of innovation, like incremental and stabilising innovation (Section 8.2).
Purifying innovation is the fourth innovation concept proposed in this thesis (Chapter 9). This concept is inspired by the idea of purifying selection in biology, which removes detrimental genetic variations from a population. Based on this analogy, purifying innovation is the process of replacing, discontinuing, or rejecting features, components, or standards that no longer contribute positively to a product’s function or appeal. Such purification has significant economic implications, thus deserving to be considered as a type of innovation selection. While it may result in the destruction of capital invested in established technologies, purifying selection also serves societal interests by halting resource allocation towards ineffective or detrimental innovations.
The discussion culminates in Chapter 10. In conclusion, the dynamics of design evolution can be understood through a punctuated model. In nascent industries, there is a high rate of product innovation as diverse technologies and design concepts emerge. This uncertain environment fosters disruptive selection, where companies explore diverse designs to find the most successful solution. As industries mature, a dominant design emerges, shifting the focus of innovation towards refining and improving the established design. This shift from disruptive to directional/incremental innovation marks a transition to a more stable phase of industrial evolution. Eventually, industries may reach a state of technological stasis and decline, leading to the discard of established designs. Therefore, understanding the full cycle of disruptive innovation, replication, and discard is essential for a comprehensive understanding of design evolution.
Contributions to Design Theory
The intersection of design and evolutionary theory has been highlighted by numerous scholars in the past (Steadman, 1979/2008; Basalla, 1988; Hybs & Gero, 1992; Michl, 1995; Langrish, 2004; 2014; Yagou, 2005; Whyte, 2007; Ehlhardt, 2016). The first section of the literature review presents an overview of this theme within the field of design research (Section 2.2). This semi-systematic review considered the top five design research journals by impact factor, as well as seminal books on the subject. However, despite decades of discussion, our understanding of how designs evolve remains limited.
While these studies have provided valuable insights, they have not yet converged into a comprehensive theory on the evolution of technology and design. The discussion continues to be characterised by divergent perspectives regarding the nature of design evolution, specifically whether it follows a Darwinian or Lamarckian process, thus indicating an ongoing lack of consensus. Unfortunately, a significant portion of the debate has revolved around comparing the similarities and differences between artificial and biological evolution.
Furthermore, this conversation has been hindered by repetitive arguments that are neither new nor ignored by proponents of the evolutionary analogy. These arguments often focus on the absence of a mechanism for sexual recombination, genetic inheritance, and the differences observed in the “family tree” of organic species compared to that of artefacts. It is worth mentioning that as early as the 1920s, anthropologist Alfred Kroeber had already engaged in discussions regarding these points. The contenders in this discussion also fail to acknowledge that, despite fundamental differences with biological evolution, technological evolution provides enough diversity (i.e., variation), continuity (i.e., inheritance), and innovation (i.e., mutation) to support an evolutionary selection process. As a result, this unsettled debate has hindered the assimilation of evolutionary principles into design theory, despite the evolutionary character of its practice.
The issue of design evolution has been a topic of discussion for decades, as the relevant literature demonstrates. However, I believe that an evolutionary theory of design is still in its early stages, leaving ample room for further development and debate. Dismissing the evolutionary analogy based on dissimilarities would be a missed opportunity, as evolutionary theory has made significant contributions to various fields of knowledge, including anthropology, computation, economics, philosophy, linguistics, and neuroscience. Evolution is too big of a theory, too powerful in its explanatory power, to settle the discussion of technological evolution in terms of dissimilarities versus similarities.
What sets this study apart from previous research is the comparison with specific patterns of natural selection and the empirical evidence presented to substantiate such analogies. Previous research has primarily focused on broad principles of natural selection, such as variation, inheritance, and mutation. In comparison, this study advances this analogy by proposing that artefacts also evolve through specific mechanisms of selection, such as disruptive, directional, stabilising, and purifying selection.
Although insightful, previous research has predominantly relied on theoretical arguments, often lacking empirical evidence to substantiate its claims. The lack of empirical grounding has likely led critics to consider that technologies evolve “perhaps in some metaphorical sense” (Lewens, 2002, p. 200). Indeed, some evolutionary concepts have been criticised as storytelling, given the difficulty in operationalising them in research and the lack of empirical testing (Whyte, 2007). The specific details of how these mechanisms work still need further development in both conceptual and empirical terms (Breslin, 2011).
In contribution, this study is one of the few that provide empirical evidence to support the proposed analogies. The results chapters provide evidence that demonstrates the specific patterns of selection in the evolution of commercial aircraft and smartphones. Chapters 4 and 5 are further divided into subsections that correspond to each specific pattern of selection observed. Grounding these concepts in data may help convince critics that evolution is more than just a “creative metaphor”.
The evolutionary perspective brings major implications for design theory and practice. Instead of the modernist credo of disruption, the conception that design is led by gifted individuals, and the “revolutionary” claims made in product advertising, evolutionary theory stresses that designs evolve through a continuous, cumulative, and collective process. This challenges the common view of design as the work of a solitary genius and emphasises the importance of collaboration and building upon existing knowledge.
Moreover, understanding the selective pressures exerted by markets and consumers provides designers with a deeper appreciation of the competitive forces that shape design evolution. Their creative minds create an abundance of novel artefacts, from which society can make selections (Basalla, 1988). Competing manufacturers and their development teams introduce a surplus of variation, comprised of many competing designs (Ehlhardt, 2016). Upon this variation, selection can occur, which means that some designs will be adopted and diffused, while others will be rejected and discontinued. This competitive dimension can not be ignored in design theory and education.
Products are often advertised as “revolutionary” or “radical,” when in truth, they are usually just small incremental improvements over prior generations. “Everything made now,” Kubler says, “is either a replica or a variant of something made a little time ago and so on back without break to the first morning of human time” (Kubler, 1962/2008). All ideas have predecessors and are always based on previous work, sometimes through refinement, sometimes through a novel combination of several pre-existing ideas (Norman & Verganti, 2014). The continuous character of technological evolution corresponds, partially, to the evolutionary mechanisms of heredity (Steadman, 1979/2008; Hybs & Gero, 1992) or retention (Ehlhardt, 2016; Eger & Ehlhardt, 2018).
Continuity is central to the process of product generation as it involves a collective and cumulative dimension (Yagou, 2005). Designers build new designs upon the achievements of previous generations, thus accumulating knowledge and skills (Steadman, 1979/2008). This collaborative and decentralised process emphasises that design evolution is much less dependent on the contributions of individual geniuses. Instead, it is the result of an organised system where each generation builds upon the work of their predecessors (Steadman, 1979/2008).
Therefore, the evolutionary perspective highlights the importance of recognising the contributions of countless individuals, including both producers and consumers, in shaping the trajectory of design. The invention of new products is better understood as the culmination of inventive work from multiple individuals, drawing from diverse knowledge domains and influenced by selective pressures from various contextual forces (Ehlhardt, 2016). This perspective encourages a rethinking of design history, injecting a sense of humility into the often inflated world of contemporary design (Yagou, 2005).
The evolutionary biologist Theodosius Dobzhansky once wrote that “nothing in biology makes sense except in the light of evolution”. Since Darwin, evolutionary theory has been, and still is, the most powerful and meaningful explanation for the diversity of life and the unity of living things. Evolution provides a fundamental and organising framework for the study of nature. It is the ligand that consolidates our knowledge, from the smallest systems (cells, genes, viruses) to the wider and more complex ones (ecology, kingdoms, biomes, life, etc). In contrast, the study of designed things has no such ligand. An evolutionary framework could help establish connections between various fields of study, including management, innovation, entrepreneurship, economics, and more. Of course, this study is not the definitive answer to this problem, but it is a small contribution in this direction.
Contributions to Disruptive Innovation Research
Another area for contribution to theory is the debate over the concept of disruption. In this work, I attempted to shed light on the neglected past of disruptive innovation research, which was largely headed by William Abernathy in the early 1980s. Research on disruption reemerged in the early 2000s, following the success of the book The Innovator’s Dilemma. The definition of disruption that Christensen proposed in this book, however, marked a significant divergence from the work of earlier innovation theorists. Since then, Christensen’s unorthodoxy has established itself as a new paradigm.
Christensen was hardly the founder of disruptive innovation research, a concept frequently discussed in the works of Abernathy and coauthors (Abernathy et al., 1983; Abernathy and Clark, 1985). In fact, Christensen does not claim to have discovered this concept, but he also fails to acknowledge the contributions of his predecessors (e.g., Christensen et al., 2018). The main problem is that Christensen appropriated and redefined the concept, disconnecting it from its research tradition.
Christensen set a new paradigm of research by incorporating a range of analytical tools, premises, and specific conditions into the definition of this phenomenon. For decades, several scholars have raised objections to this narrow definition, with no effect. These scholars have raised concerns about Christensen’s emphasis on intersecting performance trajectories, the conflation of business model strategies with the definition of disruption, the anomalies posed by high-end disruptive innovations, and the lack of simplicity and parsimony in his theoretical framework (Section 6.2).
According to Thomas Kuhn, knowledge can be lost during paradigm shifts, particularly when successful explanations from the previous theory are not retained in the new one. Unfortunately, this appears to be the case with research on disruptive innovation. According to Christensen, his predecessors were primarily focused on describing and explaining technological changes. Disruption has since evolved from this “essentially descriptive and relatively limited in scope” framework into a “more broadly explanatory causal theory of innovation and competitive response” (Christensen et al., 2018). However, how can a theory expand its scope with the introduction of numerous premises and conditions that severely limit its context and applicability?
As seen in this thesis, the new theory can no longer explain major historical events such as the Jet Age and the emergence of smartphones. This loss of explanatory power is evident in the reclassification of the jet engine as a sustaining innovation (Section 6.6). In contrast, previous scholars readily recognised the disruptive impact of the jet engine, classifying it as a technological discontinuity (Foster, 1986), a competence-destroying technology (Tushman & Anderson, 1986), an architectural innovation (Henderson & Clark, 1990), or a new dominant design (Anderson & Tushman, 1990). Despite their age, these “essentially descriptive and relatively limited” frameworks continue to yield valid explanations for major historical and technological transformations, such as the emergence of smartphones (Section 6.4), the rise of Boeing and Airbus (Sections 6.10 and 6.11), and the ongoing transition to electric vehicles (Section 6.12). It is contradictory to claim that disruption has expanded in scope and explanatory power when, in reality, the new theory no longer applies nor explains these profound cycles of industrial change. According to Kuhn’s theory, this may signify a loss of explanatory power.
In my view, Christensen’s concept of disruption seems to be a grave deterioration from his predecessors, even from his own previous works, such as his PhD thesis. In biology, disruption and diversification are synonyms. Disruptive selection leads to more diversification as it breaks the genetic equilibrium of species. This notion resounds with the understanding of pioneering scholars, who recognised that instead of “enhancing and strengthening” the industrial equilibrium, some innovations “disrupt and destroy” (Abernathy & Clark, 1985).
Not content with this simple but elegant definition, Christensen turned disruption into a highly sophisticated model. The new theory incorporates five core premises, which can be broken down into twelve sub-premises (Tellis, 2006). The analysis of performance trajectories, which is another analytical tool introduced by Christensen, is what signals the onset of disruption. Given such complexity, it is not surprising that over the past few decades “the theory’s core concepts have been widely misunderstood and its basic tenets frequently misapplied” (Christensen et al., 2015).
This study also questions the validity of Christensen’s predictive model by analysing the stock performance of companies involved in his projections (Sections 6.8 and 6.9). It highlights that despite theoretical predictions that the iPhone would fail, Apple’s stock has risen by more than 5,800% since its introduction. Similarly, in 2004, Christensen alerted that Boeing and Airbus faced a disruptive threat from regional jet manufacturers, namely Bombardier and Embraer. Instead of being disrupted, Boeing and Airbus have outperformed Bombardier and Embraer by a large margin, with their stocks rising by 800% and 600%, respectively. Even in the face of such flawed predictions, which were orders of magnitude wrong, the 2016 edition of The Innovator’s Dilemma claims that “the theory of disruption continues to yield predictions that are quite accurate, in an astounding range of industries” (Christensen, 2016).
As noted in the Literature Review, parallels between technology, design, and evolutionary theory are a recurring theme. Many seminal studies incorporate evolutionary models and concepts, specially Gould-Eldredge’s theory of punctuated evolution. This study, however, is the first to propose a framework of innovations based on specific patterns of natural selection, namely disruptive, directional, stabilising, and purifying selection. This analogy was supported by comparing these patterns described in the literature to actual data on the evolution of smartphones and commercial aircraft. In summary, technological evolution can be broken down into four primary categories of innovation selection: disruptive, directional, stabilising, and purifying innovation. These patterns generate a punctuated model of industrial evolution that, by extension, integrates Abernathy and Utterback’s model and many of their original insights (Chapter 10).
This evolutionary perspective challenges the modernist belief in disruption as the primary driver of innovation. This research has demonstrated that disruption is just one of the patterns of evolutionary selection shaping the evolution of industries. The widespread focus on disruptive innovation has created a bias towards it, neglecting other forms of innovation such as incremental and stabilising innovation (Section 8.2). However, the reality does not support this bias. As studies of technological innovation in industry have shown, companies are more interested in introducing minor changes into existing products than experimenting with spectacular new concepts (Yagou, 2005). Indeed, most companies do not expect development teams to come up with high levels of product innovation or disruptiveness:
This is understandable since most new product designs are actually new iterations of proven product concepts. For well-established product concepts, the job of the system designer is greatly simplified. A well-established product platform may be in place, so that the designer need only be concerned about adding one or two new features, leaving the rest of the design unchanged (Haskell, 2004, p. 33).
This understanding helps to dispel the notion that disruptive innovation is the sole measure of success in product development. The categorisation proposed in this study highlights that the evolution of industries does not fit into a binary dichotomy between disruptive and sustaining innovations, as defended by Christensen. Instead, the evolution of industries involves a more complex interplay of different evolutionary patterns. As hypothesised long ago, “the type of innovation that is likely to succeed, whether technologically complex or simple, and whether applied to product or process, also depends upon the stage of [industrial] development” (Utterback & Abernathy, 1975). However, given the persistent bias towards disruption, how can practitioners choose an innovation strategy that better fits with the current evolutionary stage of their industry? To overcome this bias, it is important to recognise that disruptions emerge along other patterns, such as stability and decline.
In conclusion, the research on disruptive innovation could greatly benefit from the reintegration of alternative paradigms that have faded away over time, especially the work led by Abernathy and colleagues, whom I consider to be the forgotten pioneers of disruptive innovation research. In this thesis, I argue that Abernathy’s disruption model is superior to Christensen’s since it addresses many of the inconsistencies and problems introduced in The Innovator’s Dilemma. Based on the analogy with disruptive selection, I also propose a new definition for disruptive innovations and, by extension, a new explanation for this phenomenon:
Disruptive innovation is characterised by the introduction of a new qualitative trait (like a new material, component, feature, or form factor) into an existing architecture (like a product, system, or service), or the conception of a new architecture.
That is, while incremental innovations involve enhancing existing qualities and features, disruptive innovation operates by introducing new elements into existing products or, more rarely, by introducing completely new architectures. This conceptualisation aligns with both Darwinian evolutionary theory and the pioneering work of William J. Abernathy.
11.2 Value for Practice
When it comes to practice, the constant presence of the biological theme suggests the significance of an evolutionary theory of technological change and design. Practitioners frequently employ biological terminology, such as “fitness,” “survival,” “niche,” “hybrid,” and “genealogy,” when discussing technological concepts. The parallels are so obvious to them that these analogies are often employed without the need for additional clarification (Ziman, 2000). However, given the lack of empirical data substantiating these analogies, many critics have considered evolution nothing more than just a “creative metaphor”. The fundamental differences between the organic and the artificial world have hindered scholars from considering a Darwinian perspective, in which evolution occurs through variation and selective retention. In light of these difficulties, certain scholars have shown a preference for Lamarckism (e.g., Steadman 2005; 1979/2008; Langrish, 2004). This has led some scholars to argue that variation and selection could be misleading when applied to design practice:
It is not clear what varies and what is selected, whether it is knowledge, designs, sub-assemblies, or other contenders. The practices of expert designers are not easily explained through processes for creating surplus variation; creating competition between variants, and then selecting the most appropriate (Whyte, 2007, p. 54).
The evolution of designs is a process inherently driven by selection, as evidenced by data. Selection operates at various levels and sub-levels, encompassing not only product architectures but also components, configurations, materials, and more. At any given time, multiple levels of analysis are engaged in selection, resulting in some parts experiencing disruption while others undergo incremental improvements (directional innovation) or complete elimination (purifying innovation). This layered analysis reflects the complex nature of living organisms and artificial products.
In addition, professional designers do indeed generate “surplus variation” upon which selection can occur. Sir Jonathan Ive, Apple’s former Chief Design Officer, once said that Apple’s design process is about making “lots and lots of models and prototypes” (Compton, 2017). During his career at Apple, the assembly and scoring of models and prototypes was the standard practice (Kahney, 2013). Besides generating variation, designers are responsible for selecting outsourced materials and components, which allows them to exert selective pressure on suppliers. Therefore, professional designers occupy a privileged position in technological evolution, as they both generate and sort through variation, selecting the best new traits and consolidating the optimal ones. This internal selection, however, is just the beginning. Once designers close in on a solution, the design will be evaluated by an ever larger group of people:
Many different variants are put on the market (or published, or practised, or adjudicated, or invested in, etc. as the case may be). There they are subjected to severe selection, by customers and other users (or competing groups, courts of appeal, banks, and so on). The entities that survive are replicated, diffuse through the population and become the predominant type (Ziman, 2000, p. 4).
However, there is more to evolution than just survival of the fittest. An evolutionary theory of design stresses that cooperation is just as important as competition. Products do indeed compete against each other, but such competition is also based on cooperation. The ultimate selectors (the users) are searching for the solutions that create more value for them. The designs that strike this harmony will be the ones to be selected, replicated, and retained.
11.3 Further Research
Despite a general agreement that designs evolve in an evolutionary fashion, this issue remains open for further research. This thesis has brought empirical evidence highlighting the potential analogies between the design of things and the evolution of life. I believe that evolution is more than just an inspiration or a metaphor; evolution and design may share some specific mechanisms. Further research activities could explore the external validity of such evolutionary mechanisms, grounding them in data and therefore contributing to the development of an evolutionary theory of design.
To facilitate future research, I decided to share the main datasets generated during this study in the online version of this thesis (Research Data). Data sharing practices could benefit the study of technological evolution and innovation by enabling other scholars to conduct secondary analyses, replicate, verify, and extend original findings. Considering the extent of evolutionary theory and the complexity of its models, it is inevitable that there are more analytical aspects to uncover than those covered in this study. Data sharing would advance this field by enabling alternative interpretations of existing data, thus reducing the need for data collection and fostering further theoretical debates.
More replication would help to establish the general validity and usefulness of evolutionary theory in the context of design (Ehlhardt, 2016). The patterns of natural selection (i.e., disruptive selection, directional selection, stabilising selection, and purifying selection) deserve additional research through verification in other industries. To the best of my knowledge, this thesis is the first to investigate whether these natural selection mechanisms are transferable to the evolution of things. Aside from just replicating findings, further research could refine and extend the innovation concepts proposed here.
Besides the issue of design evolution, this thesis also aims to contribute to disruptive innovation theory. Since the dawn of academic research on disruptive innovation, several scholars have questioned its very own definition (Acee, 2001; Govindarajan & Kopalle, 2006; Markides, 2006; Sood & Tellis, 2011; Govindarajan et al., 2011). Regrettably, no significant improvements have been made thus far. This thesis proposes a new concept of disruption. It re-grounds disruptive innovation on data, evolutionary theory, and earlier research traditions on disruption (the works of William Abernathy). It is far from a definitive answer. There is still much debate to be held among practitioners and academics before a final answer can be provided.
To summarise, this thesis could lead to future studies on the evolution of designs and disruptive innovations. The patterns of natural selection explored in this study (disruptive selection, directional selection, stabilising selection, and purifying selection) may become working hypotheses in subsequent research. If replicated and accepted, these concepts of design innovation will clarify the current black box of how designs evolve.
11.4 Limitations
The limitations of this study arise from two main areas: the grounded theory method and the census data. After conducting a census of two industries and drawing comparisons with evolutionary theory, the thesis presents four innovation concepts that together form an evolutionary model of design evolution. The results chapter presents an extensive list of instances, indicating the recurrence of these evolutionary patterns in two “most different” industries. However, despite the apparent objectivity of such data, it should be recognised that the human element involved in the interpretation process introduces a degree of subjectivity. Like any human endeavour, the theory-building process is inherently open to subjectivities. Despite the realist goals of Classic Grounded Theory, the researcher’s interpretation of this reality is inherently influenced by their personal experiences, analytical skills, theoretical orientations, and biases.
The second limitation stems from the census data. A census, by definition, endeavours to gather information from all eligible elements within a population. However, it is doubtful whether any census has ever successfully captured all elements, due to factors such as deficiencies in the framing process, census procedures, respondent cooperation, among other issues (Cantwell, 2008).
Despite these challenges, this study compares favourably with a study conducted by The International Council on Clean Transportation (ICCT). As outlined in Section 4.1, the census reported in this thesis accounted for a population size 3.5% smaller than the one reported by the ICCT. Considering that the ICCT study covered planes with a seating capacity exceeding 20, whereas this study focused on aircraft with a minimum of 30 seats, it was reasonable to anticipate some deviation in the overall results.
While the results from a census typically do not suffer from sampling errors, censuses are still susceptible to the non-sampling errors found in sample surveys (Cantwell, 2008). Incomplete data is an example. In this study, older records posed challenges in obtaining complete item responses for aircraft and smartphone specifications. Triangulation of sources effectively resolved these issues.
Other potential sources of non-sampling errors can be attributed to issues of coverage, in which the target universe may be over-represented or under-represented (Cantwell, 2008). In this sense, the census of commercial aircraft in this study has achieved coverage of 96.5% compared to the ICCT census. However, it is important to note that there are different inclusion criteria between the studies, so this discrepancy does not necessarily imply under-coverage in this study. As for the census of smartphone specifications, no comparable censuses or consolidated databases are available for reference. Therefore, gauging its coverage is more challenging since the total population of smartphone specifications remains indeterminate.
Last updated