Old Disks and Old Metaphors

Call it a passion project. The past few days in my spare time at work I’ve been recovering data from twenty-five and thirty-year old floppy disks. The files on these old disks—CAD drawings, meeting minutes, reports, and other construction-related documents structured in 1.44 MB or smaller bundles—are interminably boring, but there is something intellectually thrilling in the process of accessing and reviewing them. I’ve been thinking of this as an archival thrill, similar in the little raised neurons it tickles to the feeling I get when chasing leads in old newspapers or digging through a box of original documents in search of names, clues, faces. Entire careers have come and gone since these files were copied to the magnetic circles in their little plastic cases. Whole computing paradigms have risen and fallen in that time, and, with them, our own sense of technical superiority to the people who authored these files. Still, the same meticulous attention to detail is evident in the files, the same sense of their own sophistication on the part of the authors, the same workaday problems we are solving today.

Working the files, I noticed two more things:

  1. The sound of a physical device reading data is special, and it can be deeply satisfying. I had forgotten the audible experience of computing—the whining, clicking, tapping, and whirring which used to characterize the entire experience. All of this is gone now, replaced by the sterile sound of fans, maybe, like wind blowing over a dried lakebed. There are audible affordances in physical media. When the sound stops, for example, the transfer is finished. When the button clicks on a cassette tape, the experience is complete.
  2. The old files on these disks are authored with maximum efficiency in mind. With only a few hundred KBs to work with, designers had to get creative in ways we don’t today. There are a lot of pointillistic graphics, tiny GIFS, plaintext, line drawings; none of the giant, full-resolution graphics we include everywhere today.

One of the disks contains a full website, preserved like a museum piece from 1999. Clicking around those old pages got me thinking about the archival thrill of the old internet.

Consider the way that the most prominent metaphors of the web have shifted over time.

It used to be that people would surf information on the internet, riding a flow state wave across documents and domains in pursuit of greater knowledge, entertaining tidbits, or occult truths previously hidden in books, microfilm, periodicals, letters, and other texts. The oceanic internet held out the sort of thrill you feel when wandering among the stacks of a vast library or perusing the Sufi bookstalls of old Timbuktu. It was an archival thrill, tinged with participatory mystique, abounding with secrets.

In the heady days of the early web, to surf was to thrill in the freedom of information itself.

When Google arrived on the scene and began its ongoing project of organizing the information on the web, feeding took the place of surfing. This act, like every triumph of industrial capital, relied first upon the extraction of surplus value from the laborers who produced the commodity—i.e., the authors of the information. That is a subject for another day. More to my point in today’s rumination, however, Google’s revolutionary commodification of the web also took advantage of the customer’s innate narcissism. You have specific and important information needs, Google says with its design language, which this text bar can satisfy.

Google delivered on this promise by surfing the web on behalf of searchers. To deploy another (very stretched) oceanic metaphor, Google turned surfers into consumers of tuna fish. Each search serves up a little can of tuna. Enter a term in the box and out pops a little tin; pop the can and get what you need, increasingly on the first page; and then get on with Your Busy Life.

The Your Busy Life warrant is the play on narcissism. You don’t have time to surf, it says, because you are important. Have this can of tuna instead.

I love tuna. I search every day. Google was so successful, however, that the web wrapped itself around the tuna-dispensing search box. By the mid-2000s, users no longer used search primarily as an entry point to the waves but, rather, as a sort of information vending machine serving up content from Google’s trusted searches.

Beginning around 2008, feeding completely overtook surfing as the dominant user metaphor of the web. As Infinite-scroll apps on smartphones took the place of websites, the purveyors of these apps took it upon themselves to predict what users would like to know, see, or do. To this end, the most talented software engineers in the world have spent more than two decades now building algorithms designed to settle users in a stationary location and serve them little morsels of information on an infinite conveyor belt. Cans of tuna became Kibbles and Bytes, piece by piece, scrolling past.

The participatory mystique, or archival thrill, as I have called it, has been almost completely displaced by this dull feedlot experience. I know that the old experience of the web exists alongside the new, that I could go surfing right now if the urge carried me away, but I lament that so many of the people who could be building more and better websites are building cans of tuna for the Google vending machine on the web or Kibbles and Bytes for the apps.

Think of what we could have.

It Finally Happened: Bing (and DuckDuckGo) Surpasses Google

Here’s a thing I didn’t expect to see today. A Bing search just outperformed a Google search in relevance. I was looking for a relatively obscure book to reference for an essay I’m writing, and behold!

The Bing results. Note the Archive.org link. The book is available to checkout in PDF or ePub format at this link. The informational window at right is useful, as well. I was logged into my Microsoft account when I performed this search.
The Google results. With the exception of a blatant scam link in second place, all of these links point to places where I can buy the book. The first link leads to a review of the book in a scholarly journal. The Archive.org link is nowhere to be found. Notably, none of these links lead to a PDF version of the book, which was one of the search terms. I was logged into my Google account when I performed this search.

I’ve been reading anecdotes about deteriorating Google search quality, and I now I have one of my very own to share.

Update: I wondered what would happen if I tried the same search using DuckDuckGo and the results are actually better than Google, too.

The DuckDuckGo results. Note the Archive.org link in first place. A link to a dissertation hosted by my university’s scholarly commons is ranked second place. This made me suspicious, since DuckDuckGo prioritizes privacy and none of the other browsers included geographically relevant results, but the engine retrieved the same results when I ran the search through a VPN on a clean cache. Unfortunately the dissertation, though a PDF, was only tangentially relevant to the search. Still, the results are demonstrably better than Google’s.

None of this proves that these engines are better than Google as a daily driver, of course, but they certainly beat the behemoth in this edge case. With Apple rumored to be working on an alternative search service, too, competition may finally be coming to Mountain View.

Research of Note: Data Activism

Gutiérrez, Miren and Stefania Milan. “Playing with Data and its Consequences.” First Monday 24:1 (2019). https://doi.org/10.5210/fm.v24i1.9554

Scholars, advocates, and social critics frequently describe data as a structure of power used against citizens and the powerless online. In their article in the most recent First Monday, Miren Gutiérrez and Stefania Milan invert big data, arguing that “Citizens, activists and professionals alike embrace innovative data-related practices at the intersection of the digital and the informational, embedding data and ways of playing with data in their activities.

Data is undoubtedly used to oppress and exploit, but Gutiérrez and Milan show how it can be used to advocate for the rights of the less powerful, as well. Recent work in critical studies of neoliberalism–I’m thinking about Byung-Chul Han’s ideas, in particular, which are very nicely summarized in the Verso essay collection, Psychopolitics–paints a nearly hopeless picture of privacy in the radically transparent world that social media has wrought. While it does not occupy the same intellectual field, this research introduces a necessary critical counterpoint.

Problem 1,364,872 with Facebook and Data

So, full disclosure: I’ve deleted my Facebook account twice in the past 6 years. Last month I came back again after about six months away with a shamefaced grin. It made me sad to think about all of the people I know sharing their lives with one another, without me. If I’m unwilling to let those connections go, then I can’t opt out of Facebook.

But today I was reading this fantastic reporting in the New York Times about the company’s response to its many crises, and what troubled me is not necessarily that it knows everything about me–which it does–but that the best way to monetize that information is to zero in on the weaknesses: the points of ignorance, credulity, impulse, and reaction. It’s a vast database of personal pressure points the platform presses all day long.

Google has the same information, Apple has a lot of it, Samsung now owns a lot of my pressure points, and all of the apps, trackers, and aggregators on my phone, iPad, computers, and web browsers that don’t come from those companies know a lot about me. There’s no way to opt out and no clear way forward for any of us. Is there some combination of open source and paid platforms, along with encryption and data security practices that will save us?

A Thought on Indexing and Power

Tonight I was reading a book on indexing and abstracting–Brian O’Connor’s Explorations in Indexing and Abstracting (1996)–and had to stop and think about one of O’Connor’s guiding principles: “a subject is not an inherent element of a text.”

What do you mean? I thought. Doesn’t every text have a subject? Well, first of all, no. Many texts don’t have a subject, or they merely imply a subject, or they contribute to a composite subject. Ok. So then I wondered: isn’t this just hair-splitting? I had to think about it a bit more to realize that O’Connor’s point is bigger than that. The subject does not inhere in the text. You have to examine the text to understand its subject–or lack thereof. You have to master it.

But mastery is something else entirely; something not hair-splitting, but hair-raising. Mastery is the exercise of power.

The huge and ever-expanding power of technology companies in our lives today underlines an enduring cliché: information is power. But raw information is useless. Real power rests in the hands of information brokers: those who can master texts well enough to make the information they contain accessible to those who need it. This is the most significant reason why Google is so powerful. We all desperately want and need the information on the internet, but no one can sort through it all. We need a tool, therefore, that can match the questions in our heads to the answers other people have written down, and Google has done it better than anyone else.

I mention Google and the power of tech companies only to point to my insight from tonight’s reading. As a historian, I have learned all too well that both recording and interpreting information are vital instruments of power. Indexing, abstracting, describing, and organizing are instruments of power, too, which connect the recorders with the interpreters. Because the subject is not inherent in the text, to describe it is to power over both the writer, who may not have meant what the indexer says they did, and the reader, who may not ever be able to understand what the author meant because of the indexer’s choice.

This is an awesome power. It should not be taken lightly.