Tuesday, 25 August 2015

Of technical images, with Vilém Flusser

In essence, technical images represent the end of writing, in at least two senses. They are its polar opposites; and they put an end to the domination of writing as such. This is, in essence, Flusser's theory. To better understand it, one needs to see what’s specific to each of the two forms of expression/notation. And that’s where I’m going now.


Source: Miguel Frias

I've already discussed, last week and to a lesser extent the week before, that to Flusser writing is historical in the sense of its generating historical consciousness by means of linearity. With writing, the World (that terrible chaos that presents itself as our constant embarrassment) is returned to us ordered. But ordered in a special way. Ordered like the files and ranks of an army, like the wires between two telegraph poles, like the trajectory of an arrow about to hit its target. In other words, like the lines of a written text. Although I placed them first, the metaphors preceding this last sentence are the results of writing, not its models. Writing allows us to see the straightness of all these trajectories because, with writing, we have become accustomed to the paradigm of the line. Understanding, reasoning, logic, historical consciousness, etc. etc. – all these are manifestations of straightness.
The least we can say about this way of ordering is that it has transformed the World as well as our understanding of it.
But writing is not exempt from the tests of dialectics. Its rise to dominance implies some kind of fall as well. The fall may not be fully visible now, when we’re at the very beginning of a major transformation. It may never turn out to be a complete, deep, catastrophic fall. But we can’t turn a blind eye to the fact that writing is taking new forms, and that, more than anything else, it is losing ground. Of course, letting go of writing is not an easy task. Writing has been (and still is) one of our closest companions. We have built civilizations based on it. We have glorified and tortured, eradicated and constructed, invited and enforced – with writing on the topmast.

Apparatuses of transformation

But if letting go must happen, in the name of what can we be said to be letting go of writing? Flusser proposes technical images. Technical images, unlike writing, are not linear. The page of a book reads from top left corner to bottom right; or from top right to bottom left; horizontally or vertically. No matter which script is used, the principle is the same: we go about it in a straight line.
This, however, doesn’t apply to images. Images are meant to be taken as a whole. There is no linear reading of a photograph. One doesn’t start from the top left corner and proceeds to arrive at the bottom right corner.
Unlike writing, which is concerned with lines, images are concerned with surfaces. And this is a fundamental difference. But a difference that’s not so striking if we think that, in fact, representation has been making use of images for a long time. Since way before the emergence of writing, to be more precise. The caves of Lascaux are there to prove it.
So the question arises: what’s so special about technical images? What sets them apart from other visual artefacts?
Flusser focuses on one type of image: photography.

Source: Lilip Studio

Now, of course, the immediate thought crossing one’s mind in relation to photography must be something about its indexical value. Indexicality means that a photograph points at an actual object in a way similar to how the index finger does it; and it does so more prominently than, say, painting. A photographer takes a picture of something. Something that exists exactly the way it appears in the picture that’s been taken. Painting, which is also a representation of something, transforms the object. The work of the artist is apparent in every brushstroke, in every conscious use of perspective, of shadows, of composition in general. Of course, a photographer (and a filmmaker more so!) can do all this him/herself, and with relative ease. So this is not where the fundamental difference between painting and photography lies. To Flusser, the actual difference resides in the fact that the camera is a coded apparatus. Realistic as it may seem (a snippet of reality, as the cliché goes), a photograph is the result of the operations inherent in the camera. Images taken by a photographer will be dependent on the mechanical (and more recently, digital) processes made possible by the camera they’re using.
One may argue that something similar happens in the case of painting; that painting, like photography, requires certain material preconditions in order to exist. Yes, but painting preserves the human factor. Errors in painting are completely due to the artist’s application of the material preconditions. Too much colour here, mistaken application of shadows there, there’s a plethora of possibilities where a painter can go wrong. In photography (where the human remains relevant, no doubt), a large proportion of the possible errors are due to the range of operations built into the camera, and over which the user has no control. An experienced photographer will be able to apply the right apertures to the right photographic situations; but they will not be able to overcome the fact that the camera has only this many types of aperture.
So here’s the fundamental rupture. When we use a photographic camera it is not the apparatus becoming an extension of us (which is the classic understanding of technology at large) but us becoming an extension of it. When we press the button of a camera we enable the coded possibilities built into the apparatus to come to fruition. We serve the camera. We help it come to the realization of its potential, of the specific possibilities extant in its code. And that doesn’t quite happen in the case of painting.

With algorithms we jump into post-humanity

With this in mind it’s becoming easier, I think, to see how the latest optical technologies (analogue or digital) are forms of this general subjection to an apparatus. I don’t mean this in a dystopian sense, as the robot that takes over; but rather in the sense of a development whereby the human has vacated its own creation, like a Deus Otiosus that will only return, if ever, in order to punish the independence given to his creation. (Remember the biblical story of the fall?)
Algorithms are a crucial illustration of all of the above; an example Flusser did not have the chance to discuss (he died in 1991, so all of the 21st-century technologies we’re using nowadays were unknown to him; although he had a fairly good understanding of how an algorithm works).
Digital algorithms take the coding of the camera to a whole new level. They don’t even need our hand to press their buttons. There’s no button for algorithms. The functions of a button are, once again, scripted into the code. And so, apart from the initial turning-on, an algorithm needs no further human input. An algorithm is built to work in its own terms, after its own script, automated to process information (data) by applying ad infinitum the functions written in its code.

Source: Dark Government

The difference between the code of a camera and the code of an algorithm is simple: while with the former we used the apparatus ignoring (not noticing) the code but hoping to be able to control the results, with the latter we are fully aware of the code but no longer in control of the results. What’s more, we are also fully aware that there’s nothing we can do in order to stop the algorithm. Nothing but the ultimate solutions: shutting it down.
So mechanical cameras and digital codes have this thing in common: the repositioning of the human element. The human is no longer at the center of production of artefacts but at the periphery. That’s why ‘post-humanism’ is such a catchphrase nowadays.
The post-human is – due to no simple coincidence – also a post-writing. Of course, writing hasn’t been completely eliminated from the picture (my pun!). We are still deeply immersed in the linearity that’s been guiding our consciousness for thousands of years. Binary code, which stands at the foundation of the digital world, is itself arranged according to the logic of the line. Not only is it written in lines that start in the top left corner and advance towards the bottom right corner; it is also constructed in accordance to a linear causality between code and operation: code is teleologically written, so as to lead to a result visible in a computer operation.
Flusser has pointed out the fact that no definite separation from linear writing is possible as long as we keep thinking and organizing ourselves according to the principles of linearity. I’m not even sure that we should be aiming towards a complete annihilation. That kind of radicalism would verge dangerously on suicide. But major changes are taking place. Of course they can’t be sudden and complete. But they are here, they are questioning the grounds of writing. Two examples of these changes, in which the algorithm reigns supreme, are Google Earth and soft cinema. I’ll discuss them next week. So we’ll see.

Tuesday, 18 August 2015

A brief history of writing

Last week I touched quickly, insufficiently, on the issue of algorithmic reality. And that brought to my mind Vilém Flusser’s concern with technical images. And so I thought I must say a few things about that now. But before algorithms can come to the foreground I need to say a few things about writing à la Flusser. In what follows I’ll be citing especially “Letters of the Alphabet,” an essay in the collection Does Writing Have a Future?


Source: Te Ipu Pakore: The Broken Vessel
Vilém Flusser’s theories of writing have a strong historical grounding. He proposed that writing makes history. That, on the one hand, it is a practice with a background, with a beginning (Sumerian tablets, Egyptian papyri, Roman wax tablets and stone inscriptions, and so on and so forth). On the other hand, thought, writing makes History. It was only after the invention of writing that historical consciousness was made possible.

The lines that put an end to a babble

The possibility to arrange events in a linear manner, to speak of them as ordered, is the by-product of writing itself. None of this was possible before writing, when, says Flusser, the world existed in a state of mythical consciousness; when language itself was not yet settled, not yet set in stone (or in whatever).
“It is possible to claim that people of that time babbled.”
Because of the absence of writing, pre-literate cultures could not have spoken the way we do, because they had not yet gone down writing’s path.
With writing things became different in the world and in language. With writing we start to see events through the lens of their inclination towards linear ordering. One after another, events partake in a curious pageant that leaves traces on surfaces, i.e. creates a history of the event’s presence and its advancement towards the next event. With writing, we can imagine what has caused an event and what is likely to be inferred from it. That’s because we can see the before and the after of the event in a linear arrangement. And once we can see that we can’t see otherwise: events must have causes and must produce effects. Writing has made it possible to speak of origins and of projections into the future.
Logic itself is the result of writing, says Flusser. Logic, as a form of syllogistic bargaining with data based on causes and effects, on premises and conclusions, on inferences and injunctions, with the intention of arriving at a truth with a chronology of its own, is a derivative, again, of writing, and of its ability to put things into straight lines.

A simple formula

The success of writing appears to have been due to the simplicity of its formula. Once the world is presented in straight lines, it becomes easier to re-present. The simplification that came with the creation of specialized signs (letters able to synthetize the world through simple combinations) turned the human mind away from pictographic representations, which, realistic as they may have been, were time-consuming and sedentary. The Lascaux paintings are still in Lascaux because the caves could not be transported anywhere else; and so, in order to have access to the signs represented there, one had to be there, in Lascaux.

Source: Musées Royaux d'Art et d'Histoire
I can imagine elitism catching roots right there, in the darkness of those caves. Those who could not afford the privilege of seeing the murals were left outside of knowledge, outside of humanhood, forced to gang up with other prehistoric ignoramuses. Deep inside the caves, those paintings were guarded from the outside world by their very remoteness and, by being guarded, they were also preserved.
Writing did not eliminate preservation from the scheme. Hence its success. Since the ability to eternalize an event through pictographic signs was a privilege of representation per se, writing could not ignore it. So it too advanced the promise of eternity. But on top of that, communication by means of these specialized signs presented this huge advantage of being movable. The human animal turns now its attention to lighter surfaces, easier to transport: tree bark, shells, animal hides. Things found in nature (simple, sympathetic to inscription, at hand) are now turned into support for writing.

The personalization of representation

The story goes on, but there’s no time here to go into details. The point, as Flusser formulated it, is this: before writing, the connection with the world was not completely severed. Of course, the animal painted on a cave wall is not the animal itself but a representation of it: a reminder that such things exist out there, beyond the threshold of the cave. The representation, though, has full referential power: it does refer to an actual animal out there, it is an image of that animal.
Writing ruins this certainty. Writing intervenes between image and human being to unsettle their marriage. Letters are representations not of the world but of images of the world, images akin to the Lascaux paintings. This is apparent in the formal resemblance between Western letters and their referents:
“In the fifteen hundred years since their invention, their original form has changed repeatedly, and yet it remains recognizable: the two horns of the Semitic steer (Hebrew: aleph) in the A, the two domes of the Semitic house (Hebrew: beth) in the B, the hump of the Semitic camel (Hebrew: gimul) in the C. Letters are pictures of a cultural scene as it was perceived by those who invented the alphabet in the second millennium B.C. on the eastern Mediterranean.”
This original resemblance has been forgotten, but it must be taken into consideration when we think of how the letters of the alphabet (or of other writing systems) have been allowed to act as representations of representations; of how they were welcomed at the representational games.

Source: Women of Faith
And so, with the emergence of writing, the world took a step back in its relationship to the human element. Now, in order to make the cave painting accessible to others, a writer refers to that representation in an abstract way. He/she does not explain the world to their readers. What they explain is the representation they have of that world. So writing becomes not only a way of objectifying the world but also of personalizing representation, making it the product of the whim of a writer or other. Precisely because a written text can circulate, it can be present where the reality it refers to is not. A writer can describe in writing a bull or a wild horse, and the reader has to take his words for granted. In essence, there’s no way of ascertaining the truth of an utterance when the referent is absent. That’s how rhetoric became necessary: a way of persuading an audience devoid of direct access to the object. That’s how logic turned up: as a way of proving the truth of the world by means of permutations of thoughts and by establishing a diagrammatic proof-building methodology. Both rhetoric and logic deal with abstractions, with their problems as well as their solutions.

No turning back for language

And hence the essential perversion of the relationship between language and writing:
“Something in the spoken language itself calls out to be fixed in place – and in fact, not so much in the memories of speakers and hearers, or in records or tapes, but rather in the writing itself. Spoken language seems to rush toward writing almost on its own, to become a written language and so to achieve its full maturity. After the invention of writing, spoken language appears to be a preparation for a written language, to teach people how to speak properly in the first place.”
This progression mentioned by Flusser is due, of course, to the several stages of evolution that language has undertaken ever since the invention of writing. As writing turned to be language’s technical way of materialization (its own technology), the invention of inscription, once acknowledged, could never be dis-acknowledged. And so language started takin shapes dictated not by its internal forces but by the forces of its technological apparatus (writing).
“Today we have hardly any access to preliterate speech. Even in nurseries and among illiterates, writing has permeated the language.”
So the schism is total. The specialized signs of writing have taken over the realm of representation. But this victory lasted only this many centuries. Flusser finds photography to be the great rupture in the history of writing, the way towards a different form of formulating the world. With cameras and their encoded capabilities, the door was opened to what Flusser called technical images.
Of them (algorithms et al), next week, if the gods of writing are so inclined as to give us a chance.

Monday, 10 August 2015

Being caught reddit-handed, or the pleasures of policing for free

I got in trouble at the weekend when I set up a reddit account (my first one ever) and, having had a rather cursory reading through the rules of the group (or subreddit) I had joined, I was kicked out. I had, it seems, gravely violated the rule no-links-to-your-personal-blog-are-allowed-on-this-subreddit. Mind you, the link was inside a post that was the equivalent of half a Microsoft Word page, but what does that matter? Out I went. First infringement, red card. Harsh policy (zero tolerance); excellent policy-implementation.


Source: Fear and Loathing in Bioethics
The episode has left me thinking. Not about dis-affiliation from groups but about this idea of being spotted, of being handcuffed (metaphorically speaking), of being caught red-handed.
So Id like to start by noticing this language of apprehension and its potential for rhyme. Reddit, red-handed. Just for fun. (Couldnt help it.) Also couldnt help thinking about the way reddit is organized, as a means of dissemination and dismissal at the same time. But I need to be well understood: by reddit I mean pretty much all of social media, and with it pretty much everything that takes place online.

Nobody has asked for it

Now allow me a few words about the avatar that caught me. I checked his/her/their history as one does in such circumstances, Im sure. They havent posted anything original in one year. Ever since their last post theyve been engaged in replies, more or else, but also in the territorial and emblazoning (to them) actions of policing the space in search for intruders. Now thats an interesting aspect, because nobody has given them a sheriffs badge for doing so. Nobody gives any such badge to any such vigilante on the internet. And I mean nobody, because the actual work, the real work of fending and defending, of blocking and dismissing, is done not by human beings but by algorithms. What human beings can do is alert those algorithms when they come across incriminating practices such as (I had to find out) my own. And thats exactly where the issue finds its grounding. In the fact that individuals offer their time and resources to serve algorithms. Take it as a dystopian panic if you will, although I dont mean it that way. And so the deeper question that imposes itself is this: isnt this loyalty to the algorithm somehow a manifestation of the loyalty to another kind of abstractions? To the abstractions represented by ideologies?

Algorithm, the latest materialization of control
All ideologies need materializations. They cannot dwell in the abstract because theyre supposed to address concrete social items: human beings. Yes, this itemization that were subjected to comes hand in hand with the externalization of our concerns. Stuffed with the dictates we are exposed to like some docile teddy bears, we take pleasure in controlling others, in exercising this weird, dehumanizing (algorithmizing) power: the ability to find the right places where others do wrong by a given system.
The vigilante who caught me reddit-handed had a concern larger and seemingly more important than the actual reason he had presumably joined reddit for. He was there to discuss books. Thats what the group was for, thats what the group was asking for. The rules were clear (I found out after reading them closely): you dont post unless your posts are related to book discussions. Its with this kind of adamantine restrictiveness that the issue of participation was imposed on the members. And yet, the individual who caught me (I want to insist on this word because I want, in essence, to insist on the process of being criminalized which is precisely what rule-breaking leads to) did not post a reply to my contribution. No, they chose to report me (another word of the same resonance). Content was not important. What mattered was that a forbidden object was present in a forbidden place. Im not trying to defend myself because I know that would sound pathetic; and it wouldnt even be the purpose of this very post. But heres the issue: twenty lines of text mattered less than one link half a line in length. All the rest was discussion of a book! Whats more, it was a reply to a question posted by somebody else.

But lets move on

A few weeks ago I wrote about the phenomenon of synopticism: the inversion of the panoptic gaze to the point where were no longer dealing with a figure of authority watching over and super-vising the subjects but rather with the situation where subjects watch around themselves in order to police the space of their social interaction as if they were law enforcers. Policemen without a payroll. Thats precisely what Foucault pointed out throughout his career: the disciplining of the individuals to the point where they become guarantors of the discourses functioning. There are many other illustrations of the phenomenon. Think of all the white supremacist movements and their urge to protect a presumed universality by clearing it of a presumed non-white threat.

From Django Unchained. Source: giphy
Think of the neighbor who peers into your own bedroom. Think of the entitlement to install surveillance cameras on ones own property.
This last example is a really interesting one, and possibly closer to the online vigilance Im trying to bring up in this post. In essence, a camera set up on ones property will not annul a theft. It is not the camera that deters a thief from breaking and entering but the fear of being caught on it. Not only that, but the camera is not a security object per se. It does not hinder access. Like even the loudest alarm ever, it is not a blocking device. If a thief did not know that a camera existed he or she would just move on undeterred; the theft would go as planned. At least an alarm system draws attention to its own presence when it goes off. Not a camera, though.
Then whats the point? What does the person who installed the camera gain from having installed it? Peace of mind? I doubt it. We, the outsiders, are not the only ones who know that a camera is completely incompetent against a break-in. The owner knows it too. So the camera doesnt demarcate for them a worry-free territory. What it does demarcate, though, is a space of policing. If the camera catches the intruder the owner can use the image to refer the intruder to the legal discourse. The owner, therefore, has little to gain for themselves but a lot to gain for the discourse.

Unpaid work makes the world go round

This is the externalization of concern I was talking about. The state of being-concerned does not serve the individual. He/she remains on duty but without being remunerated for it. They give away an important asset (their generosity) in order to fulfill the purposes of power.
Im familiar with the term leave without pay. It is very carefully pointed out whenever an individual transgresses certain rules concerning not-engagement in the workplace. But I havent heard any instance of being on duty without pay (voluntary work excluded), although a lot of what happens in terms of this policing of the social sphere by unemployed individuals points precisely in that direction.

Source: Huffington Post
Pay needs to be regarded not necessarily as monetary remuneration. These days, when visibility has become a currency in itself, social capital needs to be understood in terms of the benefit drawn from having been involved. In what? In anything. Anything, that is, that has significance to the discourse of visibility.
And so back to my poliziotto. I mean not to him-him but to what he represents. To recapitulate, what have we here? Someone in the service of an algorithm. Someone acting for the benefit of a site thats not his own, from which he gains very little, possibly nothing. Someone whos settled into a regulated territory and has inhaled all the fumes of the site, in such a great quantity that now theyre high on the idea of duty. Which is not even an idea anymore, but rather an internalized practice. And that, without going into any other details, is characteristic of the disciplined society were inhabiting. Nothing new in itself. Nothing outrageously unknown. Just another instantiation of an age-old propensity towards self-subjection, something Foucault spent entire books talking about. And talking, and talking.

Monday, 3 August 2015

Some texts reject me, so I write them

The most inspiring texts to me are not the ones that keep me immersed in them but the ones that refuse me, the ones that send me away.


Source: Search Engine Land
It happens this way: I read and I read and there’s nothing out of ordinary about my reading. I peruse with the mind to the text, open to its charisma, expecting the pleasures, watching for clues. But the mind that reads this way is that of a stranger. What I mean to say is ridiculously simple, beyond obvious: that through this type of reading I remind myself that I am not the author but only a reader. A stranger, indeed: someone who comes from without and whose likelihood to settle within is minimal. But some texts are more than that. Some texts fill me with that curious sentiment that I am the one who’s written them; that my reading of them right now is the reading of some draft I am in the process of editing. In these situations, I can’t stop thinking beside the actual text, ignoring it as it were, heading towards a conclusion that’s not the text’s but mine.

Drafting

A draft, always a draft. Which means, in essence, that I perceive incompleteness, imperfection, room-for-improvement. I have this knack for visualizing alternatives. When I read these texts I feel the urge to imagine how I would write them otherwise. Not better, not worse. Just differently.
A draft requires careful reading, i.e. the placing of care into the text. It requires me to care enough about the text to attempt to imagine it different. And so what happens next is this: I can no longer read. That’s because the urge is now in me, the urge of inspiration (I might call it that, for once).
I can imagine a point where there’s no more room for perusal, where continuing to read is a dangerous business. Dangerous because it can cause confusion. If I am not careful enough at this point, if I don’t pay sufficient attention, I risk unconscious plagiarism – which is the worst form of all, because it takes away the pleasure that comes with the stealing of something truly valuable. I know this because I’m familiar with those moments when one can remember with embarrassing accuracy a paragraph, a sentence, a phrase, but not their origin. When that happens I feel utterly incapacitated. My mind wants to find that place where everything happened first, and that desire is so strong that I can no longer concentrate. And so a frantic search starts, one that often leads nowhere but to exhaustion.
It’s much easier to plagiarize, I think, when you know exactly what you’re plagiarizing. It’s much harder to do it when you just happen stupidly upon a fragment you didn’t even know was in your head.

Drifting

So in those moments I can no longer read. I need to put the text aside and start my own text. I need to write because something in the original text tells me with the urgency of catastrophes that if I miss this opportunity I miss everything. And ‘everything’ is an incredibly ample concept sometimes.
My own texts are very often caused by texts I’ve been reading, and whose reading must be interrupted. Those texts, in their splendor, send me away. Away to the computer, away to the piece of paper and the pen. But also away from their substance (the texts’ substance).

Source: Red State
When I have to stop reading because I need to write there is no way back. I can no longer see the original text. Its presence panics me. I don’t want to have anything to do with it. Not anymore. I turn my back to it. I obliterate it. The fact that it exists disappoints me. And this is the very same text that caused inspiration in the first place.
Incredible, the ways of writing.

Textual determinism

It’s in moments like these that I see the agency that resides in texts, their ability to stir me into action. Not the authors. They don’t awaken me to the same awareness. I very rarely feel the need to praise an author for an affect brought up by their text. Authors are not interesting. Not in themselves. An author must be an author of something. Of a text. So the text is more imperative, more interesting. It’s what exists, what needs to be dealt with. It’s what possesses the capacity to determine my actions. Textual determinism – I might call it that.
With other texts, which are more silent, less reproachful than the ones that make me write, I have a different kind of relationship. With them I don’t. I don’t start anything, I don’t change anything, I don’t make an effort. When I read these texts I make notes on the margins. Sometimes. At other times I make no notes at all. Not even mental notes. These texts don’t ask for anything. And because of their silence I remain silent too. What I want to say is that I forget these silent texts. I forget them even while I’m in the process of reading them.
But the texts that speak to me are incredibly empowering. The very nerve to get away from them is evidence to this empowerment.

A case study

Speaking of notes. I do the following when I take notes at lectures, conferences, public speeches. (I used to do it when I was a student and I’m still doing it. Every time). I start the way everybody starts. I write down the words I hear. I give the speaker my time and space and reincarnate their words onto the page, my page. At this point I am fully occupied by the speaker’s speech. To put it differently, I follow their text. I pay tribute to their gesture, and with it I confirm their authority over the text, over the clarity of that text. I would not dare thinking of altering anything. Like a good journalist who obeys the rules of his profession and protects his sources, I strive for exactitude. Everything for a faithful rendition. Everything for loyalty. But then, all of a sudden, something happens. Suddenly, the speech I am listening to ceases to be clear. It becomes blurred. It fades slowly, until it reaches inaudibility. And then, I cannot hear it at all. Why? Because at that point I am already being forced to generate my own text.
What I think happens at that point is simple, albeit brutal. I snatch the original idea. I literally steal it, the way thieves sneak into the houses of the unaware to dispossess them of valuables. And once that idea is in my possession I run away with it.
It’s the grab, the seizing of the opportunity to write, that estranges me from the speech that keeps going on in the room, unheard by me but still alive to others.

Source: Hearts and Minds
And then I write. First, things directly related to the speech. Then gradually relevance fades. It too goes away. And so the original text gives in, and in its place comes my own text, my own speech. I end up, of course, writing things completely unrelated to the original situation. But it’s now, after having encountered and then immediately divorced the original text, that I find the right energy to write. As if the echo of the original text were contaminating me.

What do you know, this too might be some kind of disease.