I’ve been thinking about how computers read things for awhile now. I first cracked the box open when I wrote this piece, part of a collection in A New Everyday cluster on Comics and Multimodality we never get enough credit for, with that awesome open-access crank and dear friend Ernesto Priego. Lately, I’ve been revisiting this idea of how computers read because I have been flipping through Speculative Everything by Anthony Dunne and Fiona Raby. To rip the publicity blurb, Dunne and Raby imagine design as a “tool to create not only things but ideas. For them, design is a means of speculating about how things could be—to imagine possible futures.” (source) I’m not that into “possible futures,” but I do like that nod to building or making ideas. I’d push it a step further, I want to think about how to represent different ways of reading. Or, more precisely, different ways of representing the act of reading, and by extension, writing.
What I’m on about is getting behind the different processes we bring to reading–expectations, inheritances, assumptions. I want to get behind them so that, frankly, the students I confront daily who are intimidated by the act of reading, or treat it as a task with rules, and thus prescriptive “cheats”–rather than as part of a process for accessing the liberally educated mind; part of the grammar for discourse (intelligent or otherwise). If I can show students how the assumptions they bring to reading might be something that can itself be interrogated, perhaps the process of decompressing anxieties can begin. In other words, get them to think less about what they are reading and more about how they are reading and how their approach to reading is embedded in the processes of things–machines, signs, images, even social structures. But I’m getting ahead here; I’m just looking at how computers reconfigure the act of reading or, make it visible in a way that allows discussion about the process of reading–and imagining–itself.
Three Stages of Computer Reading
For those in the know, I am borrowing heavily from the Latin trivium, as in a grammar, logic, and rhetoric or, as the definition makes it: input, process, and output. I chose to use Foucault’s The Order of Things because it was lying on my desk, neglected, when on whim I set out on this task. In hindsight though, the title is wildly appropriate. I can’t speak authoritatively for the content having never read it, but the wikipedia blurb about the book would suggest that the content is wildly appropriate to this rumination I’m embarking on too–so, here’s to serendipity!
I’m using the following tools: a Makerbot Digitizer, a Makerbot Replicator (5th Generation), and the Makerbot “MakerWare” software to run the digitizer and the printer. The aim is to see how the computer digitizes and then represents that digitized object. The outcome will be to see what, if anything, this might tell us about how we approach the act of reading. Also, I’m curious to see if there is anything of value in the process that might be used to get students thinking about the assumptions they make about reading or to see how the machine mimics the act of reading as a human might do it.
Here’s a shot of the original object on the scanner bed ready to go, nothing spectacular, but there it is:
I chose this position because it most accurately reflects the position the book would be in if it were being read by a human. That is, upright, slightly open or about to be opened. I did one version where the book was lying down, spine sideways, but this seemed unfair to the machine.
Stage One: Input / Scanning
The first thing the digitizer software does is scan the object. To do so, it takes a whole bunch of pictures and stitches them together to make a 3D representation on the screen. The scanner uses a laser to guide the camera around the object and allow the computer to re-assemble the object using a kind of triangulation technique that assembles the “dots” the camera has picked up into the object itself. (After you read this you can go and check out the Wikipedia entry about how this is done, but I’m going to keep moving.) I just imagine the two lasers on the scanner as a pair of eyes, opening one after the other, to create a stereoscopic image. In other words, the lasers on the scanners are the eyes that are reading the text (or object–we’re gonna be flipping metaphors frequently here, but stay with me). The scanner bed spins around so that the camera catches all sides of the object and interprets it correctly. If you look at the scanner, it really does look a little human, with it’s two eyes and mirrored third eye.
I see this as the point at which the computer is mimicking that first layer of reading, gathering the basic information, driving for comprehension rather than interpretation. Keeping the Trivium in mind, this is the grammar stage. The machine is plotting things out, gathering information, comprehending rather than interpreting. It’s an interesting way of representing to students why they might need to read a text more than once or that there are stages to analytical thought; it doesn’t just happen.
As you can see from the image above, the computer hasn’t quite pulled everything together yet, it has merely “scanned” the object to get a rough sense of where the key points are, where it stops and starts, where it meets other planes. This is a useful way of representing to students what the first encounter with a text looks like. It’s all there, but it’s not quite all there. It still needs some processing.
Just so we know for later, the scan takes about 9 minutes.
Stage Two: Process / Rendering
The Digitizer software and the 3D community use the term “meshing” to describe what the computer is doing when it starts to really process the scan into something that looks fully formed. It’s also a great term to use when we’re thinking about how to describe that secondary read, the one where we are really synthesizing what we’ve read into something we can use.
As you can see from the image, the software has now processed the book into something that looks solid. The object has coalesced into a whole, there’s fewer gaps. Again, for me, this mimics the outcomes for that secondary read–the moment where we start to fill in the gaps and make sense of what we have read so that we can use it. Use it to bring our ideas and interpretations of what we’ve read–or how what we have read influences our thinking–out into the world, into conversations.
Its important to note here that the processed book doesn’t look exactly like the original. It has been processed, interpreted, meshed. I like the way this represents how we process reading. We don’t read things the same way others do, we don’t re-present our reading the way others do, we don’t even synthesize our reading like others. What we tend to do is have some approximate similarities so that we recognize the original, but there are differences. One way to do this using the scanner would be to scan it again and reprocess it; the results would be slightly different each time–a tick here, a line there, etc. At the same time, the result looks solid. We can’t deny that what we have here is something opaque–there’s a certain “meshed” quality to what we’re looking at.
Now, depending on your computer processing power, the processing can take a second or twenty minutes. I use a fairly competent rig, so it only took about 3 minutes. That said, there is a commentary here about how the more you know, the more skills you have, the faster you can do this kind of processing, particularly when it comes to synthesizing reading materials. Speaking metaphorically, students can grapple with how processing power is relative to how much processor you have on board. Or, the more you read, the more you know, the faster you can synthesize things. As well, a nod can be made to how much faster the synthesizing can be compared to the scanning or initial deep dive might be. Reading at the beginning is slow, but it speeds up with subsequent passes. The more times you read something, the faster you can interpret it or the easier it becomes to interpret it–this is how practice works. The more internalized you can make the object under scrutiny, the more you can put your own slant on it’s interpretation–this is why Bach can be played differently and beautifully by many different people; but Glenn Gould’s Bach is so internalized that it sounds like Glenn Gould’s Bach.
Stage Three: Output / Constructing
The final stage of this little thought exercise is actually printing out the scan of the book on a 3D printer. The Makerbot Replicator, 5th Generation, is a relatively forgiving machine that can do these kind of jobs pretty well, where resolution and finesse are not a huge concern. Not surprisingly, the third and final stage of this thinking about computers reading gambit takes the longest. The full print takes around 2.5 hours.
Watching the book emerge is interesting because the process is a lot like watching the rhetorical process develop. It’s the moment where all the disparate lines (and meshings) that began the process come together to form something real, something tangible, that takes up its place in the world. In other words, one can really see the tangible result of the reading, scanning, meshing, and see how it makes something new or, contributes to the conversation. Even as the book is being rendered on the printer, the hollowed out honeycombed sections got me thinking about how we might represent chapters or how those holes might be filled with something. What would we fill them with? Sand? And then what would that mean metaphorically–an idea or sentence is a grain of sand forever locked inside the object? I know I’m pushing it here, but I think that’s a really interesting way of getting us to think about what it means when we write something and put it in a book–that book locks the words, provides a context, but at the same time access to those contexts, meanings, interpretations is difficult. One needs a key. It’s an interesting way to represent interpretation or the interpretative act.
As well, there’s something powerful about making something, bringing something into the world that has never existed before the moment you made it. I think if students look at their work in this way, rather than as something written to satisfy a quota, or rule, or requirement, they might be more enthusiastic about what they’re undertaking.
After a couple of hours, one arrives here, the completed reinterpretation, copy, object, book:
What’s interesting is that while the finished reproduction looks like the original a bit, it’s a long way from being openly identifiable. There’s no words, no title, the shape is hard to see without the original (it could be just a box). None of the usual markers for a book are there, but for those who are taking part in the process, it is an instantly recognizable representation of a final step.
There is the sense that the text has been rendered whole again, comprehended, represented after interpretation. It’s back in the real world–processed, received. Done. Sure, it’s been disrupted, reimagined, but what it represents is a rhetoric born out of grappling with grammar and logic (in the trivium sense noted above); the input and process have an output. The output is different, but it harkens back to the original in interesting ways. It might also get us to think about the original in a new way or it might push us away from the original. What is reinforced though is that all this scanning, processing, thinking, reading, gets us somewhere. And, that somewhere it gets us makes a contribution to an ongoing conversation–it allows us to build processing power.
For the computer, the reading has resulted in an output. The student can see how her reading, scanning, processing leads to an output. Where the computer metaphorically renders a 3D model, the student might metaphorically see how she renders an essay. Point is, something has been inscribed on our world, we can now handle, manipulate, reprocess that “new” object. The inscription can be called upon, archived, remembered, it can even execute change upon how we see the original or, more bravely, ourselves. Not to overstate the case, but the way the computer reveals reading is to expose the sometimes dark nature of the processes behind thinking, learning, and responding to an ongoing discourse.
Final Thoughts and Reflections
There’s a point when you do this kind of work where you just reflect back and think: “this is so dumb; what the hell am I wasting my time on?” I certainly felt that way about this little project, especially when the write up of it looks like it’ll be just south of 3000 words.
Throughout the project, I was very aware that as I was at once observing the process of reading, interpreting, representing, I was also reading, interpreting, representing. I was very active thinking about what I was doing, the ramifications for the decisions I made. I was to some degree in partnership with the computer; we were both working at making something.
All that said, there’s lots of space to think about things within this project described above. I want to think about “mid-way interventions”; what could we do with the spaces inside the plastic 3D printed book (the filling it with sand stuff above)? Could these be thought of as chapters, a place to lock things–how does this reflect the impossibility of total reinterpretation (issues of translation might come to the fore here)? How is the completed copy a kind of “closed” form, a way of representing the impenetrability of texts in a way that doesn’t rely on the “reading is hard” or “students are lazy” models–how can these reproductions and the stages I have outlined here help with understanding teaching?
As I take this thinking forward, I will move on to getting students to think about scale, what happens when we re-scale something? What does that reveal about the original, the copy? (The completed object I’ve described here was about 1/5 the size of the original book. Printing the original size would have taken too long.) I’ll also be looking more closely at how projects such as these tell us something about the technical requirements of reproducibility, what does it mean to only use copies? Does looking at copies actually encourage innovation or, is being able to invent the new where we need to direct ourselves?
So What About This as an Open Resource?
One of the obligations we have when we are doing work in the open, making open pedagogical materials, sharing, is thinking about how we can redefine our practices in order to debunk the mythologies students often have when they enter the classroom. You know the ones: reading is super hard, writing is hard, thinking is hard, it’s all way too hard and I am just here for a C+, so tell me what I have to do to get that please and I’ll be on my way. If we can redefine our own processes by making them visible to students, who can then see and rethink for themselves what the act of reading or writing or interpretation might be, then we’ve really opened something.
In the end, the job of being open is not to teach; it is to show how the processes of learning–in context–work. Access to materials, textbooks, classes are a huge part of what open educational resources can do. There’s also a need to peel back the onion on the mythologies of what we do. To make our own process, our own trivia, more visible to students. Too often we get stuck in teaching things. Instead, let’s show, out in the open, how we are making things–be they essays, objects, ideas, or arguments.
Share this post:
Licensed under Creative Commons Attribution-NonCommercial-ShareAlike 2.5 Canada