The Nature of Technology

Introduction

Technology is such a central feature of the music sector that it is appropriate to describe a recent theory of how it works. Much of what is described below in a general context also applies to the music sector. Readers are invited to reflect on this and respond with their comments. Maybe there is a case for a specific article as suggested in the note on The Nature of Music Sector Technology?

W. Brian Arthur’s The Nature of Technology¹ is important for understanding how technologies develop, and therefore how future technologies may emerge. This article encapsulates his thesis.

Arthur distinguishes between individual technologies (“technologies-singular”) such as a jet engine or radar, and “domains” (“technologies-plural”) such as electronics, radio engineering or biotechnology. Invention in the sense of radically new technology is also obviously important. How they link up is described below, but first we must note the inspiration from the two most quoted persons in his book: Charles Darwin and Joseph Schumpeter.

Combinatorial versus Biological Evolution

Technologies evolve in the sense that they all, including the novel ones called inventions, descend in some way from the technologies that preceded them. Arthur defines technological change as a organic process, giving it a living quality. Technologies are “ecologies” with components and practices that must fit together at any time, change constantly as new elements enter, and throw off little sub-colonies from time to time that have a different quality. This provides a parallel to biological evolution, a link which according to “Arthur’s Second Law” (see the endnote of this article) will strengthen in the 21st century²

In genetic evolution, new combinations are created in incremental steps which must produce a living organism at all stages. This does not work in technology. Schumpeter said in his Theory of Economic Development (1912, 1934): “Add successively as many mail coaches as you please, you will never get a railroad thereby.” Rather, every new technology and solution is a combination of parts from preceding technologies, facilitated by the constant capture and harnessing of natural phenomena that make the technology possible. In technology, Darwinian-type variation and selection happen when technologies are adapted to practical applications through engineering design, refining already formed structures.

Arthur acknowledges Schumpeter for the thoughts he first developed a century ago in his theory of economic development. The central concept is that technological development is an endogenous force, coming from within the economic system.

Schumpeter’s theory that change in the economy arises from “new combinations of productive means” translates into “new combinations of technology” in modern terminology (Arthur 2009, p 19). Schumpeter, of course, was almost alone among economists with his dynamic views on the integral role of technology in the economy. Arthur is one of a growing band of contemporary economists who have come to a similar view (no doubt aided by his polymath background in engineering, operations research, and complexity theory): “The overall collection of technologies bootstraps itself upward from the few to the many and from the simple to the complex. We can say that technology creates itself out of itself.” (p 21)

Individual Technologies

Individual technologies are means to fulfill a human purpose. The means can be processes or devices, which are the same thing at different levels: A technology embodies a series of processes – its “software”. These operations require physical equipment to execute them – its “hardware”. If we emphasize the software, we see a process or method. If we emphasise the hardware, we see a physical device.

Individual technologies are based on three principles:

• Combination: Every technology consists of parts organised around a essential idea that allows it to work. The primary structure consists of a main assembly that carries out the base function of the technology, and supporting subsystems.
• Recursiveness: A technology consists of component building blocks that are themselves technologies, and these consist of sub-parts that are also technologies, in a repeating (recurring) pattern. These hierarchies can have several layers, all parts needing to be mutually compatible, which adds to the time and cost of developing new technologies.
• Phenomena: A technology is always based on one or more phenomena that can be exploited and used for a purpose. Capturing phenomena is the basis of all physically based technologies, and helps explain their evolution: more sophisticated phenomena such as quantum effects could not have been uncovered without the prior discovery of the electrical phenomena³.

There are actually two types of “technology” that represent means to fulfill a human purpose. Technologies such as radar or electricity generation are based on physical phenomena. But there is a huge range of other means of fulfilling human purposes such as business organisations, legal systems or contracts which are based on non-physical effects: organisational, behavioral, and even logical or mathematical procedures such as algorithms. These are also “technologies”, or at least first cousins within the class of purposed systems.

Infrastructure is a related concept. Originally referring to public works from roads and bridges to schools and hospitals, it is now defined as comprising both the physical and the organisational structure needed for a society or enterprise.⁴

Science is also a purposed system, which is vitally important for the evolutionary process because science and technology co-exist, or “co-evolve”, in a symbiotic relationship. Modern science provides the formal uncovering of new phenomena and the understanding and knowledge of these phenomena. Technology builds from harnessing the phenomena which nowadays are being almost exclusively uncovered by science, but it also supplies the instruments and methods and experiments that science needs to develop its knowledge.

Domains

As families of chemical, electrical, quantum or other phenomena are harnessed, they give rise to bodies of technologies that work naturally together (domains). These differ from individual technologies such as radar, which achieve particular purposes. A domain (“technology-plural”) is a toolbox of useful components to be drawn from a set of practices to be used. Domains are more than the sum of their individual technologies and operate in a different way from these. They determine what is possible in a given era, and what becomes the characteristic industries of an era. There is nothing static about this. What can be accomplished constantly changes as a domain evolves as it expands its base of phenomena.

Engineers play an important role as designers of individual applications requiring new combinations of components to achieve a specific purpose. The specialised designs combine to push an existing technology and its domain forward. In this way, experience with different solutions and sub-solutions steadily accumulates and technologies change and improve over time. Engineering therefore makes vitally important contributions to the evolution of technology and to future innovation.

The Origin and Development of New Technologies

How do radically novel technologies arise – technologies using new or different principles for a particular purpose (the equivalent of new biological species)? Laser printing differs in kind from line printing technology, jet propulsion from propelled aircraft engines, and computation using electronic relay circuits from computation using electro-mechanic means. The new technologies use principles that are “radically novel” – they are inventions.⁵

A radically novel technology, however, always emerges from an accumulation of previous components and functional characteristics already in place. It is the culmination of a progression of previous devices, inventions, and understandings that led to the new technology. Supporting any novel device or method is an underlying pyramid of causality: of other technologies that used the principle in question; of antecedent technologies that contributed to the solution; of supporting principles and components that made the new technology possible; of phenomena once novel that made these in turn possible; of instruments, techniques, and manufacturing processes used in the new technology; of previous craft and understanding; of the grammars of the phenomena used and of the principles employed; of the interactions among people at all levels.

Particularly important in this pyramid of causality is the accumulated scientific and technical knowledge. It is contained within engineering itself, but also within universities, learned societies, national academies of science and engineering, and published journals. These form the critical substrate from which technologies emerge. Importantly, the logical structure for invention extends to origination in science; without the accumulated knowledge and capacity to innovate in science, the basis for technology would be vastly poorer.

Typically the initial version of a novel technology is crude and must be made reliable, scaled up and applied effectively to different purposes. As a technology develops, it runs into barriers and bottlenecks that are usually overcome by replacing the impeded components with more efficient components. This again leads to requirements to adjust other parts to accommodate it – it may even require a rethink of the architecture of the technology.

In addition to internal replacement, another mechanism of development is at work, which Arthur calls structural deepening. Here the component presenting an obstacle is retained, but additional components and assemblies are added to it to work around the limitation. By adding subsystems, technologies add “depth” and design sophistication to their structures as well as complexity. The two mechanisms of development, internal replacement and structural deepening, apply through the life of a technology. Early on, a new technology is developed deliberately and experimentally. Later, it develops further as new instances of it are designed for specific purposes – it becomes part of the standard engineering process. But eventually there comes a time when neither component replacement nor structural deepening add much to performance. The technology reaches maturity. In the absence of a suitable novel principle, the old one gets locked in.

The locking-in of an older successful principle causes adaptive stretch. But at some point of development, the old principle cannot be stretched further. The way is now open for a novel principle to get a footing. The origination of a new principle, structural deepening, locking-in, and adaptive stretch, therefore have a natural cycle. Eventually the old principle is strained beyond its limits and gives way to the new one. The new base principle is simpler, but in due course it may become entangled itself.

The Development of New Domains

Many domains coalesce around a central technology. Computation spawned supporting technologies which might themselves develop into separate domains: printers, punched card readers, external memory systems, and programming languages. Other domains form around families of phenomena and the understandings and practices that go with these. Electronics and radio engineering were built on the understanding of electrons and electromagnetic waves.

A nascent field starts as part of its parent domain. Genetic engineering in its medical applications began as a minor offshoot of molecular biology and biochemistry. There is an experimental feeling about these early days; participants see themselves as solving particular problems in their parent domain. But in time, the new cluster acquires its own vocabulary and way of thinking. Eventually domains reach old age. Some perish, but most live on: we still use bridges and roads, sewer systems and electric lighting, much as we did 100 years ago (though individual technologies within the domain develop, sometimes quite considerably like electric lighting over the past few years), but we still tend to take them for granted as technologies.

Not all domains go through a cycle of youth, adulthood and old age. Some disrupt the cycle by reinventing themselves every few years. They morph when one of its key technologies undergo radical change. Electronics changed character when the transistor replaced the vacuum tube. The communications and information technology domains keep morphing.

Besides morphing, a domain evolves new domains, often with multiple parentage. Information technology including the Internet is a child of telecommunications and computation, resulting from the marriage of high-speed data transmission and high-speed data manipulation. The parent fields live on, but they have birthed things that exist on their own, and now a good deal of their energy flows to the new branch. This tendency to morph and to sire new domains is part of what gives bodies of technology a living quality.

The economy adjusts when it encounters a novel body of technology. A new domain is described by its methods, devices, understandings, and practices. A particular industry consists of its organisations and business processes, production methods and physical equipment, all widely defined as “technologies”. These two collections of individual technologies – one from the new domain, and the other from the particular industry – come together and influence one another, resulting in other new combinations.

It All Takes Time

The unfolding of the new technology and readjusting of the economy takes a great deal of time. A revolution will eventually happen when the businesses and commercial procedures of the economy are arranged around its technologies, and these technologies adapt and become competitive on general compatibility and costs with other technologies that deliver similar benefits.

For this to happen, the new domain must gather adherents and prestige, find purpose and uses, resolve obstacles⁶ and fill gaps in its components, and develop technologies that support it and bridge it to the technologies that use it. Markets must be found, and the existing structures of the economy must be adjusted to make use of the new domain. The new domain must be recognised and the engineers who command the grammar of the old need to retool themselves for the new. All this must be mediated by finance, institutions, management, and government policies, and by developing skills for the new domain.

Domains therefore evolve differently from individual technologies. Once invented, the commercial development of a jet engine is basically focused, concentrated, and rational. Domains emerge slowly around loosely understood phenomena or new technology, and build organically on the components, practices, and understandings that support these. As a new domain arrives, the economy encounters it and alters itself as a result.

Arthur says on page 181: “There is no reason that such evolution, once in motion, should end.” But the process may take decades.Brian Arthur and Wolfgang Polak in ‘The evolution of technology within a simple computer model’ (Complexity, 2006, Vol.11, No. 5: 23-31) demonstrated technological evolution in which the elements are common logic functions. Starting from a primitive technology, new circuits representing new technologies are constructed by randomly wiring together existing ones and testing the result to see whether they satisfy any existing needs. If a circuit proves useful – satisfies some need better than its competitors – it replaces the one that previously satisfied that need. In the computer model it then adds itself to the active collection of technologies and becomes available as an element for the construction of still further circuits. So elements constantly add to the set of active technologies as they find uses, and leave again if rendered obsolete by others. In this way the collection of technologies bootstraps upwards by first creating simple technologies that satisfy simple needs, then from these more complex technologies to new ones that satisfy more sophisticated needs. –

In several 250,000-step runs, the system evolved an 8-bit code, the basis of a simple calculator. Arthur in The Nature of Technology (p 184) says that this would be impossible (one might say equivalent to the legendary monkey typing Shakespeare’s plays by randomly hammering the keyboard) without the model creating a series of stepping-stone technologies first, which were then used as building blocks to create circuits of intermediate complexity to be finally used to create more complicated circuits. Arthur and Polak found that if they took away the intermediate mechanisms that called for these stepping-stone technologies, the complex needs went unfulfilled.

Endnote: Arthur’s Three Laws

The Edge World Question Center asks a different question each year which elicits 150-200 responses from prominent people with a wide range of skills and experiences. In 2004 it was “What’s your law?” W. Brian Arthur came up with three “laws” which form an interesting background to his work on technology and provides an additional perspective on the future. Their subject matter figures in The Nature of Technology, though the treatment in the book doesn’t immediately convey an impression that it should be elevated to the status of “laws”. The prompt the question gave Arthur in 2004 to come up with some basic principles has therefore proved very useful.

Arthur’s response is quoted verbatim.

Arthur’s First Law: High-tech markets are dominated 70-80% by a single player – product, company, or country. The reason: Such markets are subject to increasing returns or self-reinforcing mechanisms. Therefore an initial advantage – often bestowed by chance – leads to increasing advantage and eventually heavy market domination. (Absent government intervention, of course).

Arthur’s Second Law: As technology advances it becomes ever more biological. We are leaving an age of mechanistic, fixed-design technologies, and entering an age of metabolic, self-reorganizing technologies. In this sense, as technology becomes more advanced it becomes more organic – therefore more “biological.” Further, as biological mechanisms at the cellular and DNA levels become better understood, they become harnessed and co-opted as technologies. In this century, biology and technology will therefore intertwine.

Arthur’s Third Law: The modularization of technologies increases with the extent of the market. Just as it pays to create a specialized worker if there is sufficient volume of throughput to occupy that specialty, it pays to create a standard prefabricated assembly, or module, if its function recurs in many instances. Modularity therefore is to a technological economy what the division of labor is to a manufacturing one—it increases as the economy expands.

Author

Hans Hoegh-Guldberg. Entered into knowledge base 23 August 2011.The article is an edited version of an addendum to Hans Hoegh-Guldberg, ‘Technology and Climate Change’, Background Paper 4 to Climate Change and the Florida Keys, National Oceanic and Atmospheric Administration (NOAA) and Florida Keys National Marine Sanctuary (FKNMS), 2010. The article describes what is considered the generally applicable role of technological change. See also Background Paper 3 of the same project, ‘The Changing Economic Paradigm’, especially the section on the Santa Fe Institute, of which William Brian Arthur is a prominent member.

The Santa Fe Institute is a leader in complexity theory based on the broad interdisciplinary qualifications of its members. Arthur himself obtained an undergraduate degree in electrical engineering, followed by two MAs in operations research and mathematics, respectively. He obtained his PhD in operations research in the same year (1973) that he took an MA in economics. At 37 (1982) he became the youngest endowed professor at Stanford University, in economics and population studies. Arthur has had great influence on the development of complexity economics, and complexity theory generally, at the Santa Fe Institute with which he remains associated as an External Professor, as well as being a Visiting Researcher at the Palo Alto Research Center for commercial innovation (PARC). He was joint winner of the Schumpeter Prize in Economics in 1990, when the theme for the prize was evolutionary economics. In 2009 he was joint winner of the inaugural $110,000 Lagrange Prize in Complexity Science.