The Stuff of Bits

Everything is bits, but not all bits are equally significant.

At Los Alamos during the Manhattan Project, those building the first weapons were constrained by the material processes of uranium enrich-ment. For contemporary designers, by contrast, material constraints on design are more likely to be processing capacities of their computers. The development of new computer architectures, from vectorization and array programming in the 1980s to advanced high-speed interconnections for large, distributed architectures today, frames the feasibility and possible extent of digital simulation. Indeed, one could argue that the limits on contemporary weapons design are those imposed by digital materials, not radioactive ones.

The prevailing mythology of information, as laid out by Claude Shannon (Shannon and Weaver 1947), holds that it is an abstract entity, a property of configurations of matter, so that a signal in a wire, a flag on a pole, or an arrangement of tin cans might convey the same information, but the information remains independent of that matter. In this book, I argue that the material arrangements of information-how it is represented and how that shapes how it can be put to work-matters significantly for our experience of information and information systems.

楽しみ！！情報がどのように貯蔵されているか、その物理的なコンテナの性質が実は情報社会に大きく影響を与えているという指摘。

Walter Ong (1988), for example, examines the relationship between orality and literacy and argues that each represents not just different forms of expression but different worldviews. From the perspective of a literate culture, we might marvel at the acts of memory involved in remembering and reciting long poems or stories accurately; however, accuracy in the sense of word-for-word reproduction is not a consideration in oral societies, since there is no written record against which something can be compared.

So, while the "same" words might be expressed vocally or in writing, the expectations of what these words mean and the role they are to play in communication and coordination are quite different.

デザイナーはやっぱりメンタルモデルをデザインしていて、多くの人にとってよりよい「考え方」があるか模索している。どうしたらその営みをcapture できるんだろうな。

In The Domestication of the Savage Mind, Goody (1977), one of the first proponents of the literacy hypothesis, looks beyond the basic properties of the written word. He sets out the entwined history of Western knowl edge practices and representational forms, arguing that different representa tional forms provide unique structures for encountering, organizing, and knowing the world. Lists, hierarchies, and tables provide structures and techniques for presenting and manipulating information, and thus, for understanding the world. As such, one can see how their development as representational strategies is harnessed to the development of the different modes of knowledge production associated with them-classification and ordering, generalization and grouping, and relationality and comparison.

Besides the linearity of text, Goody argues that what can be set out on the page-with the representational capacities of different lexical devices-shapes what can be known. These lexical devices become information tech-nologies with associated material properties. While lists suggest hierarchies, columns create paths for addition and subtraction; lines and arrows enable categories and groupings. Such devices provide written texts with logics of manipulability and preconceived relations.

え、これじゃないか？表現方法によって何ができるか、何を考えられるかが規定される。この本はガチで読んだほうがいい気がする

Janet Vertesi (2014a, 2015) discusses image processing and manipulation by planetary scientists working on the Mars Exploration Rover mission. Her ethnographic work highlights the embeddedness of forms of image repre sentation within larger systems of capture, transmission, manipulation, and analysis.

人類学者はフィールドワークするとき、事前に主張したいことを考えてるんだろうか。行ってみたら全然問いと違ったみたいなこともありそうだけど。

3. Finding the material in the virtual

I want here to use the case of emu lators to examine virtualization as "rematerialization"-not as a move away from the material to create a domain of the virtual, but rather a new mate rial foundation for digital experience.

デジタルなものにもmaterialityがあるとする立場からは、エミュレータなどが行うvirtualizationはmaterialから遠ざかる行為というよりも、その上で新たなmaterialityが展開される、rematerializationとみなされる。

例えばCRTのレトロゲーム機をエミュレートするときには、ビームが上下に動く時間だとか、一瞬ビームがオフになって上にリセットする間の黒画面のときに重い計算をまとめてやるハックとか、もちろん元のCPUが持っていたバグとかにも気を遣って再現しないといけない。ソフトウェアはどこでも動くバーチャルなものだとよく言うけど、めちゃくちゃ物質的な依存関係を持つ。今動いているソフトウェアも必ず将来から見たらそうなるだろう。

For example, before speakers were built in, some of the earliest efforts in computer music were achieved by indirect means, by placing a transistor radio close to a computer and "listening in" to the electromagnetic leakage from the computer's circuitry (Doornbusch 2005). Reproducing the radio-frequency emissions of an old computer might be interesting but is unquestionably beyond the capacities of most emulators, or even of what we imagine it means to be an emulation.

トランジスタラジオを近くに置いて電磁波の漏れを使って音楽を再生してたのマジか

ソフトウェアはvirtualなものであり、ソフトウェアの上で別のソフトウェアを動かすなんてdoubly virtualに思えるが、エミュレータの事例を見るとそうはなっておらず、materialityで満たされている。むしろdoubly materialである。

What is dissolved here is the illusion of layered abstractions-the idea

of of

that software systems can be effectively built up, layer by layer, with each layer adding new functionality that depends on the layer below. The notion of layering is most widely deployed when we talk about networks-so, we talk about a cloud service being defined "on top of" HTTP, which itself is implemented "on top of" TCP, which is "on top of" IP, which runs "over" the underlying hardware communication mechanisms of the connective media. The idea is that at each layer, the lower layer is both augmented (with new functions) and hidden (so that upper "layers" of the system need not worry about the details lower down).

真の「抽象化」なんて存在するのか？

4. Spreadsheet

Our work involved more than four years of immersive team ethnogra phy with a distributed scientific collaboration oriented around a spacecraft in the outer solar system and managed by NASA.

スプレッドシートの分析をするために4年間もNASAでエスノグラフィやってる

A spreadsheet starts not blank but empty. The grid shows a structure waiting to be filled in. Once a spreadsheet contains data, it also displays a structure by which those data will be talked about, debated, and managed. （略）In other words, the grid "anticipates" — both data and action.

In this way, spreadsheets through their anticipatory properties become an important way of guiding discussion.

5. NoSQL

In what Miller (2003) calls "the virtual moment," the model of the world (in the database) suddenly drives rather than docu-ments human phenomena, so that the materialities of that model make themselves visible through the ways in which the model drives institutional action.

データベースは現実世界のモデルと言うだけでなく、その特定のデータベースシステムが持つ性質によってむしろ現実世界でのアクション・意思決定・開発可能なソフトウェアが左右されているという主張。

6. The materialities of internet routing

One Wilshireというビルの話

Verizonなどの通信事業者が大量に入っていて、めちゃくちゃ重要なIPXがある場所。NTTやAT&Tがパケット交換をするらしい。

7. Internets and Othernets

Although the development of CSNET produced many important techni-cal contributions, CSNET's most significant legacies might be historical and institutional in that, first, CSNET represented a significant expansion in the spread of TCP/IP in the academic research community, and second, it was designed in collaboration with and in order to support the mission of the National Science Foundation (rather than, say, the military backers of ARPA and ARPANET). The CSNET effort laid the foundation for a later NSF-funded network called NSFNET, and it was the NSFNET backbone that was later opened up to commercial traffic, and then in 1995 replaced entirely by private service providers.

NSFNETが今のインターネットの原型だったのか。

Nonetheless, some settings for network transmission test the limits of TCP/IP's adaptability. For instance, using TCP/IP to communicate from Earth to a rover on Mars or a more distant spacecraft poses major chal lenges. Even at the speed of light, signals can incur a round-trip delay of fifteen minutes getting to Mars and back, and a round-trip delay of three hours between Earth and, say, the Cassini probe in the Saturn system. Asa second example, consider a network connection implemented by mount-ing Wi-Fi access points on buses in rural areas. The intermittent connectiv-ity that could be provided here is, in theory, sufficient for basic network transactions (submitting a web search request this morning when the bus goes by, and waiting for a response this afternoon when it passes this way again), but these delays go well beyond what was imagined when TCP/IP was designed.

別の遅延があるネットワークの事例として、超遠隔地においてバスが通ったときだけネットが使えるようになるサービスもルワンダで提供されていた。ある日住人が調べたいことを入力すると、翌日のバスでそれがわかる。

This brief survey of alternative internetworking provides a basis for think ing more expansively about the network design space.

The goal of an analysis of representational materialities is to do lust this-to bring the specifics of technological arrangements into view as mat-ters of social and cultural concern.