VR Mapping

For the last few days, I’ve been reading about and tinkering with White and Le Cornu’s ideas of VR mapping. This is the idea that Internet users can be differentiated along continuum between those who are essentially quick-hit, one-and-done task-based users of the Internet and those who are actively socially engaged. As a contrast to Prensky’s dubious theory of digital natives vs. digital immigrants, VR mapping is a good start toward a more accurate picture of technological engagement, though I honestly find the focus on social interaction a bit limiting.

White has a good post explaining VR mapping basics, along with a video and a much longer, detail-dense paper. For my money, you’d do well to approach them in that order. Here’s the nut of the discussion:

When in Visitor mode, individuals decide on the task they wish to undertake. For example, discovering a particular piece of information online, completing the task and then going offline or moving on to another task. In Visitor mode individuals do not leave any social trace online.

In other words, Visitors use the Internet the way men stereotypically use Home Depot. In. Get what you want. Out. Residents are the proverbial horse of another color.

When in Resident mode the individual is going online to connect to, or to be with, other people. This mode is about social presence…for example, posting to the wall in Facebook, tweeting, blogging, or posting comments on blogs. This type of online behaviour leaves a persistent social trace…

So Residents are engaged with the biomass (to borrow Stephenson’s term) behind the Web. The core contrast here is “no social trace” vs. “persistent social trace”.

Finally, White proposes that most of us don’t live completely at either extreme, but rather engage along a continuum between the two, and differentiated by personal vs. professional mode (which is again a continuum). So what White proposes is essentially a graph with Visitor and Resident along the X axis, and Personal and Professional along the Y. For measuring what it does, I think this model is fine. That said, it has some shortcomings.

I sat down to graph my own continuum, and after a lot of fiddling, wound up with this (which I still regard as evolving).

At first, I piled more icons to the right side. Then I reread White’s definitions, and moved them to the left. And it was about this time I started to realize where this model falls short, which I think it does in a few ways.

My primary argument with Prensky is that the “natives” vs. “immigrants” narrative implies some sort of inborn facility with technology that I can tell you as a higher-ed technologist with nearly two decades under my belt simply does not exist. My own daughter got a dose of this objection recently when, approaching her 13th birthday, she asserted that her “generation just knows how to use technology, because we’ve had iPads and cellphones”. I replied that my generation and her grandfather’s built the hardware and the software that she has such facility with, and wondered aloud whether she could write her own cool new app if she wanted to.

Prensky’s model treats the ability to use technology as the superior of the ability to understand it, create it, and maintain it. As anyone who can drive a car but has to pay the mechanic every time the dashboard lights up, that ain’t necessarily so.

White abstracts this one level further…the VR continuum is exclusively a measure of social interaction via the Web (and to be fair, it never directly claims to be more). This is fine as a model of social engagement, but falls down when you start talking about technological engagement in a broader sense. Sure Johnny posts a ton to Facebook, and is on Snapchat with his friends every night. But he can’t tell you how to the iPad he’s using actually works, and he has no notion of how it’s built. Susie can use every new app with tremendous facility, and actually makes money from her Instagram and YouTube feeds, but couldn’t begin to explain the human cost of our technological society.

That’s not to say White model is valueless…it’s not. But it is one-dimensional, and does nothing to encourage any deeper understanding of what it means not just to engage with technology, but to grapple with its implications. To return to the car analogy, for decades we’ve reconfigured our cities and living patterns to accommodate cars, without thinking about the longer-term social and economic impact (local, national, and global) of that shift, and now that chicken is coming home to roost. The progressive ubiquitization of technology, absent a thoughtful evaluation of the impacts of that process and a generalized willingness to engage with its consequences in meaningful ways, is going to produce a similar day of reckoning.

All of which probably seems like an unnecessarily dark train of thought to emerge from a fairly simple attempt to chart and categorize social interaction via the web, but it’s really more of a generalized objection to our ongoing failure to engage technology at a deep and thoughtful level. White’s chart is interesting, but it’s a gloss. Leaving aside for a moment concerns about social interaction vs. coding and hardware, I still think the model is a debatable metric as a basis for understanding digital identity.

The chart looks only at engagement vs. use, and personal vs. professional. Again, it’s interesting, but it doesn’t attempt to address important aspects of digital identity. My own digital ID isn’t just my years-long Twitter stream or Facebook feed (and it’s worth noting here that although those are viewed as interactive services, the bulk of my tweets and posts generate minimal interactivity…just being billed as “social” doesn’t magically make a service interactive, or its users social butterflies). My ID is deeper than that: banking record, Amazon purchases and wish lists, Dropbox storage, Flickr albums, Reddit “recent links” lists, and so on. As big data becomes more of a factor in our daily lives, understanding the true depth, breadth and composition of our digital identities is crucial.

I’ll also point out that this model is valid only insofar as users are willing to be truthful. Frequenters of online hate sites, consumers of digital pornography, participants in conspiracy theory boards, users of  online substance abuse or depression support groups, and similar individuals with habits they’d rather keep quiet are going to present a picture of their digital identities that’s incomplete, at best.

So if we want to use White’s model as a starting point for evaluation, discussion, and understanding, I think it’s fine. I just question how far it really takes us. (I also wish it offered a way to track longitudinal changes in profile, but that’s really outside its scope, and not a fair criticism.)

2 Comments on “VR Mapping

  1. Thanks for this very thoughtful post. Interesting thoughts on the layers of “digital” identity. I agree that engagement is one aspect and understanding is another (important) layer. Understanding shapes how one uses and engages. I would push back though and say even those with little understanding of how things technically work can have a robust digital identity. For me, digital identity is what we can see of someone. If the understanding of how things work is important to an identity, I would ask, where do we see that? Where and how is it communicated? To go back to the your car analogy, if I am not a mechanic by trade, but I know how to change the oil in my car, is this part of my identity? It wouldn’t be part of how people saw me, unless it was something I thought was important to tell them.

  2. “say even those with little understanding of how things technically work can have a robust digital identity”

    “digital identity is what we can see of someone”

    I think we have to take a deeper, more holistic view of identity. If I say to you “someone stole my identity”, that implies a trespass very different from, say, hijacking my Twitter feed. We spend a lot of time telling people to be very careful what they say, do, and share online, but if we restrict that to an “identity” consisting solely of what’s shared with the general public (and I note here that “general public” is just one of many publics or audiences we share with), then we leave gaping holes in folks’ understanding of what constitutes identity and what needs protecting.

    To me, what White is talking about isn’t identity, but persona.

    “Harrington proposes that we create a taxonomy for identity management. Although he doesn’t refer to the Tower of Babel, the allusion is present when he says: “today we have numerous organizations and vendors preaching different terminology…which just leads to confusion on the part of the user (and the marketing operations of the vendors). This needs clarification.” As a start towards developing this taxonomy, Ed suggests a specific hierarchical relationship for the terms “identity,” “persona” and role.” As he explains it:

    At the top is Identity – “A constant that cannot change. In the past it was validated by my fingerprints, today it is validated by my DNA.”

    Next is Persona – “An application of my identity to a broad situation – my office persona, my parenting persona or whatever. It applies my identity to a specific situation. This is often referred to as a “role,” but I think this is incorrect in that a persona may have multiple roles.”

    Next, then, would be Role – “A specific application within a persona. In my office persona, I may have a manager role, a mentor role, an employee role, etc. In my parenting persona I may be a disciplinarian or possibly a “buddy,” and so on.”

    Turn this around, so that “Identity” is at the bottom, and it’s almost an object oriented description with each persona inheriting characteristics from its various roles and the Identity being the sum of the multiple personas its derived from.

    These definitions do need some work, but for now I’ll use them in this way whenever we talk about these concepts. You should talk amongst yourselves to try to refine these relationships and definitions. But do keep me informed and I’ll spread the word to everyone else.” [http://www.networkworld.com/article/2336210/access-control/defining-identity–persona–role.html]

Leave a Reply

Your email address will not be published. Required fields are marked *

css.php