De-situating Spectrum: Non-spatial metaphors for wireless communication

Today’s spectrum policy debate is conducted in terms of a pervasive metaphor: that spectrum is spatial, signals are phenomena in space, and radios are agents (see my DYSPAN 2007 paper “Imagining Radio: Mental Models of Wireless Communication”, PDF).

Metaphors are not right or wrong, but a dominant metaphor can forestall new insights. Any metaphor highlights some aspects of the situation it is explaining, and suppresses others. We are so used to thinking about spectrum in spatial terms that the metaphor’s disadvantages have become invisible, and alternatives to it are barely thinkable. Spectrum allocation is imagined in terms of property rights in real estate, and cognitive radios are conceived of as smart agents acting within delimited, space-like frequency domains. Alternative metaphors can stimulate new approaches for policy making and research by de-familiarizing wireless communications.

It is useful for a new metaphor to have some overlaps with existing ways of thinking. An entirely different conceptual structure that shares no referents or mappings with the dominant view – like the “wireless communications are like the Internet” metaphor proposed by Open Spectrum advocates – will be so alien that it has trouble getting traction. This paper therefore explores conceptual models for wireless communication that have similar concept-to-concept mappings as the spectrum-as-space analogy, but that do not rely on a spatial metaphor.

One possible alternative metaphor is to imagine spectrum to be like all the crafts practiced in a medieval city. A spectrum license is like the letters patent that give a particular guild a monopoly in their respective crafts. The spectrum licensee is the guild. License-exempt spectrum is that set of trade activities that do not require letters patent. A spectrum part corresponds to a specific craft. Signals are like the practice of the craft. Radios are like craftspeople and their customers. The regulator corresponds to the King or city burgesses who issue charters.

One can model different types of allocation as different terms on the business practice monopoly that is issued to a craft guild. Command-and-control spectrum allocation would be analogous to the state not only issuing a charter to (say) the weavers’ guild to make textiles, but also specifying the kinds of fabric they may make, and the prices they can charge. They need the Crown’s permission to change any of the parameters, and weavers have to stay weavers. The analogy of transferable, flexible-use allocation would be that weavers still need letters patent to practice their craft, but can make any kind of fabric they like, charge any price, and can sell on their practice to anyone they wish.

In short, this is a “spectrum-as-business-practices” model. Other non-spatial metaphors include spectrum-as-intellectual-property-rights, and spectrum-as-markets (as distinct from markets in spectrum).

A new metaphor highlights hidden assumptions in the dominant conceptual model by emphasizing and downplaying issues differently. For example, the spectrum-as-business-practices analogy blurs the distinction between the craft of a weaver and the act of making and selling textiles; that is, the distinction between spectrum and signal is played down. By comparison, the dominant spectrum-as-space metaphor emphasizes that spectrum exists independently from signals, which are “added” to the space. It implies that one can clearly separate the underlying asset (spectrum) from what you do with it (the signals), just as land is distinct from the buildings on it. The business practices metaphor therefore calls into question the meaning of the term “spectrum resource”, and the entire project of maximizing the utility of a nation’s spectrum resources, since spectrum is not an asset independent of use.

The spectrum-as-business-practices metaphor also stresses that the primary regulatory consideration is permission to perform a set of behaviors within certain constraints; the use of a resource, which is primary in the spectrum-as-land conception, is downplayed. A business license is a set of rules of permitted behavior, rather than use-neutral title to an asset conceived as a space-like frequency interval. It therefore invites a formulation of wireless communication policy purely in terms of signals and radios, with no privileging of frequency above other operational parameters.

Through these and other comparisons, non-spatial metaphors for wireless communications mark out an alternative conceptual model for spectrum policy and cognitive radio research that can lead to new solutions.

Indoctrinated against the Market – really?

Stefan Theil claims that French and German students are being taught that capitalism, free markets, and entrepreneurship are savage, unhealthy, and immoral. This is a crucial observation; such a pity, then, that the proof he presents in his Foreign Policy article Europe’s Philosophy of Failure is little more than a litany of inflammatory citations.

One can buttress almost any argument by selective quotation from a large enough corpus, and Theil does a good job. (Another good job is this video clip making the case that Top Gun is a gay love story.) However, a claim as sweeping as this requires at the very least a quantification of the frequency of critiques of the free market relative to neutral and positive statements. Even better would be opinion surveys of students that establish which biases they’ve absorbed; and best would be to show a correlation between specific texts/curricula and such opinions. He does quote polls, but they’re just ones of entire national populations.

I would not expect such rigor in Newsweek, where Theil is European economics editor. The readership of Foreign Policy, however, is well able to absorb a more academic approach – and deserves one.

It’s even more important to provide solid evidence when an assertion squares with most Atlantic intellectuals’ existing biases. (It definitely does with mine.) If he’s right, the divergence between the US and the EU will only increase over time, to both sides’ detriment. The implications for diplomacy and business strategy are profound. Even a greater loss, then, that he doesn’t give us better evidence to go on.

Gardening the Web (2)

I’ve been working away at a new way to think about internet/web policy – one of the reasons why this blog has been so quiet for the last few months. I hinted at this work back in November in the post Gardening the Web.

I’ve now written up a draft; I’d appreciate any feedback. If you’re eager, try the 60-page paper (DOC, PDF); there’s also a 12-page precis (DOC, PDF). For everybody else, here’s the abstract:

Technology and business shifts constantly change the context in which regulators have to make policy for the internet/web. There are also abiding policy imperatives that always need to be met: revenue, economic vitality, public safety, consumer protection, and culture and values.

Policy makers use mental models to frame policy questions and actions. However, technical change calls into question the approaches used successfully to date. This paper outlines a schema for communications regulation that reflects the nature of the internet/web. The framework is inspired by the metaphor of the garden, and is grounded in an understanding of complex adaptive social systems, of which gardens are one example.

A consideration of systems theory yield tactics (tools) and strategies (principles) that guide policy making in meeting the over-arching regulatory goals of stability and productivity. The three guiding principles are fostering experimentation, designing flexible policies, and building in resilience. These principles are implemented through tools such as encouraging diversity, keeping an open mind, delegating, setting clear boundaries, taking a holistic approach, transparency, and modeling policy choices.

The application of the framework is illustrated in a discussion of three current policy questions: video regulation, rules for internet voice services, and licensed vs. unlicensed spectrum allocation.