Automated anthropologist (some thoughts)

Roughly a week has passed since my experiment in San Francisco and some thoughts are in order.

For those who are new to the enterprise: On July 16, I installed myself in SF and invited people to send me instructions via Twitter.  I promised to do pretty much anything people asked me to do.  

It was a disaster.  And not in an interesting way.  But in a “how could he get this so wrong” way.  

My plan was to be truly automated, to do in real time whatever I was instructed to do.  If someone said, “turn right” that was what I wanted to do, assuming that it did not put me in the path of an oncoming trolley.  If someone said, “burst into tears and wait for someone to come to your aid” I intended to do that too.  (More on motives and objectives in a moment.)

I failed at the automated thing.  The fact of the matter is I’m a nervous nelly.  So I cheated. I took assignments sent me by email with me into the day.  And then I asked my assistant Maria to decide which of the tweets received we would act on.  (Maria Elmqvist is just graduating from the Academy of Art University.  I had written to Cameron Maddux there to see if he knew of a student who could help out.  Maria volunteered).  This too destroyed the randomizing quality of the undertaking.  (Again, more on the point in a moment.)

In the press of the moment, old habits prevailed.  I have done a lot of ethnographic interviews in the street.  And before I knew it, I was interviewing people.  This created some interesting moments as when it become clear that a would-be respondent had just told me indirectly ‘to fuck you and leave me alone.’  Then the media found us, and that lead us to Jonathan Bloom, a really interesting guy who works for ABC7 in San Francisco.  We started chatting and it turns out that Bloom is helping reinvent the world of TV journalism and I wanted to find out more about that.  Then he started driving us from place to place.  And by this time, my head was spinning and I was thinking, “So why did I decide to do this, again?”

So why did I decide to do this?

First, Automatic anthropologist was a culturematic and every culturematic is a hack of culture.  It creates an event designed to engage, provoke, reveal culture.  

In this case, turning yourself over to the direction of other people might be expected to raise questions about agency and autonomy.FN1  Specifically, “Who’s in charge?”  And “How can someone surrender control of the self to other parties?” 

The Automated anthropologist was designed in haste.  Suddenly, I had a free day in SF and I thought, “now what?”  I am just finishing a project for the Ford Foundation in which the question of individualism surfaces almost constantly.  So I was thinking about autonomy and what it is to be a free standing individual.

As Americans we are deeply devoted to the idea that we are in charge.  We make choices. We craft lives.  We are self inventing.  The idea of voluntarily giving up this agency and autonomy strikes us as odd. (And to the media, it turns out, irresistible.) Outside of S&M dungeons and other romantic encounters, giving up control is actually unamerican.  We define ourselves by the idea that we are self defining.  

The fact of the matter is we are only partly choosing, in charge and self inventing.  We are deeply constrained and defined by social rules, cultural meanings, political forces and economic realities. I don’t make too much of this.  I am not one of those social scientists who think that because we are sometimes determined by forces outside ourselves, we are wholly defined by them.  Choice makes an extraordinary role in American life. But there are moments, ghoulish, quite scary moments, when we glimpse the limits of our autonomy and I wonder if the automated anthropologist could become one of these.

More simply, I think some people heard about the automatic anthropologist and thought, “Great.  A monkey on a string!”  It was as if they had wandered by and discovered that someone had left the door to selfhood wide open, with the keys still in the ignition!  And they had an “evil genius” moment.  

“Ah ha!  My agency will inhabit his agency.  I will make him do things that embarrass him.  I will force him to hold himself up to ridicule.  Finally, my chance to play the puppet master!” Americans are deeply opportunistic (I mean this in the technical sense) and this looked like one hell of an opportunity.    

A higher objective of the undertaking was magic.  Culturematics at their best have a way of “reenchanting the world,” to use Max Weber’s phrase. In place of the rational, the routine and the routinized, they are designed as a way to make something wonderful happen. This is what I’d been hoping for.

Perhaps the most compelling objective of the exercise was novelty, creativity, innovation, to pile up the words we use so often these days.  One of the paths to innovation is randomness.  And we see a passion for this these days in our passion for improv and experiment.  And the Automated Anthropologist looked like a way to use randomness to march me out of the world I knew into a world I didn’t.  We are self defining.  We are captives of our own little gravitational fields.  

These fields are the proverbial “boxes” we are always claiming to be trying to get out of, but it’s hard.  Many of our choices have hardened into habits.  It is very hard to escape ourselves and I thought that automation and the real time feel of advice from others might walk me straight out of the world I construct for myself into something new.  (We talk grandly and often about empathy, but this is, in my opinion, merely a matter of letting difference into consciousness on a day pass with an armed guard.  The chance of assumption-rocking transformation is remote.)  

The learning, then, is clear.  If you are going to do an event like this, you have to be scrupulous and disciplined.  You have to stick to the plan.  And you have to follow it wherever it takes you.  No cheating.  And that means you can’t do any of your own documentation. Leave that to someone else.  Your job is to be completely automated…by others…all the time.  

The learning may also be “don’t sent a nervous nelly on a mission like this.”  Or maybe that’s just a note of personal criticism.  

A note of thanks.

Sometime in the 1990s, while living in the Danforth neighborhood in Toronto, on Saturday mornings, I would wander up the record store near the Danforth subway station and fell into conversation with Dave Dyment there who introduced me to the art of the Fluxus movement and Yves Klein (see Leap, pictured).  I would not have undertaken the Automated Anthropologist without this instruction.  


FN1. Cliche alert. I blanche a little when I write this. How many exhibit catalogues have told us that the artist is “dealing with the whole question of agency.”  (Plug “whole question of agency” into Google to see what I mean.)  This has become a kind of boilerplate, the thing curators says about art without saying anything more about the topic, thus betraying reflexive behavior at the moment they wish to be critical.  With some powerful exceptions of course.  


For the Storify summary of the event, have a look here.  

For the book from which the project stems, have a look here.  

8 thoughts on “Automated anthropologist (some thoughts)

  1. Rick Liebling

    Ok, so, you gave something a shot, it didn’t turn out exactly as you expected/hoped and you learned a little something. Now, let’s look at the alternative had you not done this. Another lunch or coffee with someone? Been there, done that. Browse a book store (they still probably have those in SF).

    All things considered, I’d say you came out ahead.

    1. Grant Post author

      Rick, oh, absolutely, I’m glad I did it, and I may well do it again. Grant

  2. Indy

    I think for me, one lesson was “technology is close to making this kind of thing possible but frankly it isn’t there yet” (this kind of thing being sharing an autoanthro experience in real time).

    Ways it could be worked:

    1) Wait for Google Glass…

    2) Follow instructions off Twitter and leave it to your backup team to document – maybe real time publishing, maybe after the fact.
    Turn left when asked to or whatever, but otherwise just plow on – don’t interact back with the twitter stream.
    (Maybe assert some robotic principles – “if left unguided for 20mins AutoAnthro will locate the nearest coffeebar and order a macchiato”)

    3) Don’t even document, just write it up afterwards? Use your own device for notes/photos as memory aids only.

    I had some other thoughts about the project more generally, but I’ll have to dig them out.

    1. Grant Post author

      Indy, excellent thoughts, all. The tech is so close to real time reporting and if I understand Google Glass correctly it reports directly to the Cloud and could serve as a glass bottom boat for anyone who wanted to follow along. And yes, not reporting back would is the more scrupulous approach. Indeed, forgetting altogether about the fact that anyone is observing would be wisest. Robotic principles is a good idea and the way to solve “dead air” moments.
      Very useful. Thanks! Grant

  3. Jared

    Disaster or not, I really enjoyed the few videos of your experience that I came across. The ex-Harley employee? Wow.

    1. Grant Post author

      Jared, yes, Craig immediately disrupted the usual street person stereotypes, didn’t he? And I really should spend more time doing interviews in the street. a deindustrialized, artisanal approach. thanks, Grant

  4. Lloyd Davis

    Grant, I thought you rocked. And I think you’re being unduly hard on yourself here, but we can put that down to your honorary semi-British status.

    I was amazed at the number of pranking tweets you got – as an obvserver, I thought this was the most interesting aspect of the exercise. Should you have given yourself over to being a performing monkey? I don’t think so, you were being an Automated *Anthropologist* not simply an automaton. But yeah, you learned something and we all learned it with you.

    For time-difference reasons, I didn’t watch you all day/night, but for me I loved the bit where you filmed the interchange with the woman who’d just been grabbed in the street. I was also intrigues by the use of glympse, I’ve never seen it in real-time before, it felt a bit creepy. Well done all round.

    I think the beauty of culturematics is that there’s an immediate result/output/whatever, but then you can add to the raw data with reflection and analysis when you’ve got some distance. That’s what I’m currently doing with my two coast-to-coast American trips – they were great in themselves, but writing the stories over again, drawing on my blog and twitter and flickr and youtube is creating something much richer and meaningful.

    Good work, keep it up, I’m proud of you!

    1. Grant McCracken

      Lloyd, very grateful for your kind words! yes, I got the feeling that I had my transatlantic hat on while writing this and wondered when I will ever be fully socialized as an American. That’s the moment I will stop apologizing which is seen to be bad form in these parts. As painful when present as when absent in the UK. Looking forward ward to your summing up of your real road trips. Thanks. Best, Grant

Comments are closed.