Part I: Thinking through tabs
Our own bodies and minds allow us to experience the world, but we are also inexorably bound by our bodies. Our bodies of experience — the apparatus of ‘us’ — are both our mediator and our prison. In a fundamental sense, we are unable to extend our physical and mental physiology into our environment.
In some distant future, we likely won’t be so bound by our biology. Our minds and bodies will likely be deconstructed and reconstructed among the universe. It’s difficult to imagine a scenario where this isn’t the case, in fact. Transhumanists have long felt this way.
“Humans+”, the transhumanist idea that we will be more than human
They see a a future in which our minds, enhanced by computers, biology, and artificial intelligence will be scarcely recognisable. Our bodies, formerly structurally bound by an epidermal layer, will be porous, extending and encompassing appendages of our choosing.
But I submit — as do many a philosopher — that this future is already here. Our cognition already extends beyond the barrier of our skull.
The evidence for this is abundant, as evidenced by something that is perhaps just a glance away from you now: browser tabs.
Browsers tabs have long been seen as a useful way of collocating informational experiences such that they are easily accessed. A browser’s tabs lowers the amount of activity required to locate a pre-existing information source and negates the need to end engagement with a particular information source. Of course, windows previously have had this ability, but tabs are less distributed across the computer ecosystem — they are more immediate representations of information artifacts.
But in the digital climate we find ourselves in, tabs also act as what is known as external cognition or computational offloading. What these terms collectively indicate is a method for using external representations to reduce the amount of cognitive effort required by a particular agent — usually a person. Essentially, using this definition, tabs are more than just a method to easily re-access information, they act as reminders for what you were doing.
Notes to oneself are the most typical examples of external cognition.
Like writing on a sticky note, putting your keys near the door so you won’t forget them, or highlighting some text of importance, external cognition relies on the environment to help you cognate.
But the extended mind thesis takes this a step further. It would say that:
Tabs act as processes that form cognitive feedback loops based on on epistemic action.
What in god’s name do I mean by that?
It’s a well established that we take incomplete pictures of information — when we glance at a wallpaper full of pictures of identical Marilyn Monroes, we don’t encode every Marilyn Monroe — a full composite picture — rather we satisfice to get an overall understanding of what is being represented. We know that the wallpaper is “a series of images of Marilyn Monroe”; we don’t take a high resolution image in our mind. If we were asked to recall the wallpaper and examine it in our mind to determine which Marilyn was different, we’d fail miserably.
You’re not a camera, you don’t take high resolution pictures.
To fill the gaps in these incomplete pictures, we use what’s known as epistemic action to get information “just in time”. Epistemic (i.e. about knowledge or its validation) action is the act of manipulating that which will help us with the mental activity of a task. It is distinguished from pragmatic action, which is the activity that is the completing of the task. Turning a puzzle piece around to check and see if the shape will fit (rather than turning it around in our mind) is an epistemic action. Placing the piece in the puzzle is a pragmatic action.
Put another way: we use epistemic action as a way of by transforming the structures in our environment so we can sample them. So, if we are doing our taxes, we might have some papers around us, because it’s easier to glance at a paper with a number on it than to remember the numbers, especially if there are many of them. I might glance at a piece of paper, or move one closer to me, or I might highlight a transaction I am uncertain of. I’ll also glance at the wallpaper of Marilyn Monroes to remember which one is different.
But another principle — the principle of ecological assembly — states that we recruit resources to cognitively sample only and as justifiably as necessary. So lets take the example of browser tabs — sometimes it may do just to look at a tab heading to remember pertinent information that that tab contains, or what that tab represents. At other times, it may require actually clicking the tab to get the information required. Both of these; looking, and looking and clicking, are epistemic activities that involve sampling the environment to get just enough of what information is required. We need just what is required- just enough to fill in our incomplete picture. (Incidentally, the requirement just to look at the tabs headings to recall the pertinent information within requires an effective form of semiosis on the part of both the tab and the interpreter — see my article on semiotics in HCI).
But is information you haven’t read on tabs part of your extended mind? Well, what matters is that you know what that tabs contain and you endorse it as effectively true. I have used this example in a previous article, but imagine Otto who has Alzheimer’s and can’t remember where the Museum is. However, he carries a notebook around with him all the time and knows that the information is in his notebook. In this case, we’d say that the notebook is part of his extended mind as he knows — or believes — the whereabouts of the museum is in there. Were it in his head or in his book, the process of retrieving the information is functionally the same (that is, it performs the same function), if not technically.
So, with tabs, I might create an environment (let’s called it by its proper academic name — “cognitive niche”) where each tab relates to something I am thinking about.This cognitive niche may not necessarily be one I designed through an ontological structure (i.e. sorted by certain self-defined categories) but rather may be one created by chance. So I may have a series of tabs sequenced next to one another randomly — perhaps only defined by the chronology by which I opened each tab. But I am of the creator of this niche, and can adjust as need be.
Now, if this cognitive niche were in your head in the form of memories, what would be the difference? We sample and manipulate memories in a similar way to the way we sample and manipulate tabs. We regularly have incomplete pictures in our minds and have to consider and recall a variety of thoughts. The thesis here then is that browsers tabs are part of our mind in terms of function. That is to say, what has the information doesn’t matter, it matters that it performs a particular function for us.
But, you might say, isn’t this boundless? Surely we can say any activity or interaction with the physical world is extended cognition. Browsing a library, shopping, or even talking to people might be included. And where would it end? Wouldn’t the chain of cognition continue increasing until the entirety of the internet or even the world is part of our extended cognition?
Yet there’s a number of important factors that differentiate tabs, specifically:
Assumptions about the personal availability of information
Extremely low levels of epistemic activity result in information
Ecological assembly integration through multiple dimensions
Personal availability of information
Information in a browser tab is extremely accessible. It sits within a digital environment, it can be carried in a laptop (and in a phone, though in a different format). It is predictably accessible as well — moreso than a memory. Unlike a book or a piece of paper it can have multiple instantiations, appearing in multiple different mobile and non-mobile iterations. It’s reasonable to assume that a group of books spread around you, with bookmarks and notes in each may too act an extended mind, but relative to tabs, a group of books would be much less easily accessible, and thus it’s a lesser form of the extended mind. Tabs have a high quotient of “being at hand”, moreso than any preexisting cognitive niche.
2. Low levels of epistemic effort
Information within tabs is available with extremely low levels of effort to recover. Tab names are initially accessible through a simple saccade and fixation of the eye — less time than it takes to access most memories. This has been well studied by Ballard, who found that most people would flick their eyes to a figure than try to remember it when solving a problem involving that figure. Additionally, accessing the tab takes less than a second as well — a simple mouse movement and click. Here then, information is accessed faster than most memories. Unlike other potential forms of the extended mind — conversation, books — tabs require very little epistemic activity. However, information on these tabs requiring further clicks to access is less your extended mind as it requires more epistemic activity to recall. It’s also less likely you can functionally believe you know information that is further clicks away.
3. Ecological assembly integration through multiple dimensions
Let’s say you were doing research on Huskies — there was one at the shelter and you were wondering if it was right for you. You google Huskies and open 3 tabs:
A tab containing the Wikipedia tab about Huskies
A tab containing a webpage about Huskies’ history
A tab containing an online forum for Husky owners
You quickly read each through. You have an amalgamated cognitive niche about Huskies within your mind, but you’ve also created a feedback loop where you have each ‘container’ of information within each tab — a cognitive niche outside your mind. Information relating to the tab is accessible through the “reminder” of looking at the tab at the top of your window (which, as noted, can generate thought related to the information within the tab) and also clicking on the tab to actually read the information therein. So if you were reading about the history of Huskies on the Wikipedia page you would reflect on the knowledge you’ve created within the husky history tab or the husky forum via:
the information within your physical brain — by remembering what you’ve read
your extended mind by using epistemic activity to reference that information within the tabs (by glancing at the tab at the top of your browser or by clicking on the tab and reading that information)
This happens on multiple feedback levels. You are thinking about multiple things when you research a topic — whether you want to or not — by referencing previously instantiated information (again, either by remembering in your brain or by using epistemic activity). This information is instantiated both in your brain and in tabs, and as noted, is similarly accessible and personal.
What’s key about all 3 of these factors is that they enable your brain to expect and integrate the information available via the tabs. In this way a loop is formed between your cognition and your browser tabs.
So, we have a personal, supremely accessible, customised system of inputs looped into our cognition on a multidimensional basis. Now, you might argue that you’re certain don’t use tabs this way, and perhaps you really don’t. Or perhaps — and I’d argue this is much more likely — you do in a way that this simply isn’t apparent to you.
Remember, the nature of our brains makes it such that being aware of our cognition is in fact unhelpful when we don’t infact intend to “meta-cognate”. In other words, self-reflection is useful, but as Heidegger noted, when we are using a tool we aren’t reflecting on that tool, we are focusing on our goal. We only focus on the tool, or in this case the tab, when something goes wrong.
So it’s very likely that you aren’t aware of your extended mind because it happens through effective unconscious operation.
But, in the end, isn’t this all just a trick of language? Why does it matter what we decide is part of our mind or is not part of our mind?
Were we to consider interfaces/information as part of our cognition, this would free us from the user-tool based conceptual restrictions, allowing us to conceive of new and more effective ways to actually think. For example, if we were to think of tabs not as just browser functions but as cognitive feedback loops we would perceive their utility and hence the design of them, much differently.
Imagine if hovering a mouse cursor over a tab over-layed the current page with that tab’s page until the mouse was moved again. Or perhaps users could highlight areas of content within a tab, and that content would appear when the mouse cursor was over the tab. Or what if tabs themselves had better and perhaps customisable signs on them that allowed us to recall the information therein with more ease. These are roughly thought out examples, but they reframe our perception of how we think of tabs.
Hovering over a tab could show a relevant section of text from that tab
With the extended mind in mind (sorry), our focus would be lowering barriers to accessing information, and by making it more instantly accessible and more personalised. It’s much much easier to consider how we relate our internal and external thoughts within and between each other when we utilise the extended mind thesis.
The extended mind’s conceptual structure helps us to understand how epistemic action should be prioritised over pragmatic action in information rich environments. The ability to quickly collocate, immediately access and cross-reference information becomes of paramount importance.
Of course, the difficulty with this is that, as I’ve noted in a previous article, the structure of digital systems are metaphors or extensions of preexisting physical systems. This means that systems are not intrinsically set up to support extended minds.
In the case of the tab, its development followed from the structure of the webpage, itself a metaphor of a physical, paper page. Webpages, in essence, are a metaphor for a millennia old system of recording linear spoken language rather than something sensitive to the potentiality of new forms of cognition.
The same is true for interactive physical systems. Rather than using a new system of typing that could leave one had free to engage in epistemic action, we used the keyboard, a hangover from the typewriter, as the main interaction device with computer.
The father of HCI, Douglas Englebart invented a unique system for one handed typing that allowed the other hand to use a mouse. This would have allowed the other hand to be involved in epistemic action, but his vision died for being “too complicated”.
Douglas Englebart’s one handed “keyset” . Taken from http://web.stanford.edu
But things might slowly be changing.
Material design seems to make epistemic action important by allowing for the movement of panels of information on multiple axes.
Cross integration of multiple programs using single sign-on allows the quick access and transfer of information.
Still, we are a long ways away from what could be. And because of our familiarity with the current system, and, more importantly, our deterministic belief toward what constitutes cognition, progress is slow going.