A theory of context

The brief paper entitled A Mathematical Theory of Communication by Claude Shannon, written in 1948, is often described as the foundational document of the Internet age.  It gave us the term “bit” for the foundational unit of information (a 1 or 0), either in use by a computing process or as a piece of data in transit.  The paper is really not about communication as we think about it, that is, the process of people exchanging meaning amongst one another, either instantaneously in conversation as it may occur or across time; rather, a better title would be “A Mathematical Theory of Data Transfer in the Presence of Uncertain Errors in the Transmission Medium”.  Not nearly as catchy, of course, but far more accurate, and far less likely to confuse a reader who stumbles across it and thinks about the implications on critical theory and epistemology.

Shannon explicitly calls this out in his paper when he admits that the semantic aspects of meaning are irrelevant to the engineering problem of moving data, but the paper has had wide-ranging implications on our understanding of how meaning is moved in the electronic age.  I’m interested here in expanding a mathematical theory of data transfer to a broader conceptual framework for a theory of communications.

In the engineering problem, Shannon sets up a basic schema between a number of actors and processes:

• an information source which produces a message for transmission;
• Three engineering constructs: a transmitter, which could be a telephone or a router, a channel, the fiber optic cable or the radio signal, and a receiver, which receives the transmission and processes it back into the form in which the transmitter is believed to have broadcast the message, and finally,
• a destination, which is a user (intended or not) of the message originally created by the information source

I think this is a helpful framework, and while Shannon avoided doing so, I’d like to try exploring an extension of it to embrace the ways by which a source put stogether a message with semantic meaning and then how a destination incorporates and discerns the meaning in that message that is relevant to them.  (For the rest of this paper, I’m going to assume away the engineering and media aspects of transmitter, channel, and receiver, and we’ll assume we’re dealing with people trying to communicate either in person or via some telecommunications equipment that doesn’t meaningfully alter the audio-visual content being exchanged – although MacLuhan fans can write to me offline deriding this assumption.)  Specifically, I’d like to propose that both actors in the communications act have contexts, which we can imagine as the extensive set of assumptions, background data, and external data which inform how the information source will construct the source message, including the meaning of that message, and which will also inform (differently, which I think can be demonstrated pretty readily) how the destination will translate the transmission into intelligible information.

Modern communications theory goes a small step in this direction in its applications around cryptography.  Cryptography presupposes that the source and destination both need to share a consistent set of encoding and decoding tools which enables a message to be something other than random noise – not random noise in the sense of being just a jumble, but also in the sense of a message like “The blue moose crosses the highway with a limp”.  To the French Resistance fighter looking to understand the timing of a drop of supplies, such a message may be rich in meaning, but to the unknown Gestapo agent listening into Armed Forces Radio for encoded messages, their lack of context gives them no information whatsoever – the sentence is intelligible as an English sentence, but devoid of meaning without the context, especially since the words themselves are absurd when considered outside of any context.  This leads to a broader exploration in cryptography of how sources and destinations can exchange both information and the “codebooks” which enable their ready decryption; codebooks here are effectively a simple way of thinking about contexts as I describe it above.

What I’d argue, though, is that by elevating this notion of context as being fundamental to the nature of both the source and the destination of communication, we can see how any conversation is in fact an encrypted exchange between a source with its own context and a destination with different context.  Our “codebooks” for everyday exchange are quite complex: at a basic level it is language, with its deep structures of grammar, but it is then overlaid with societal connotations for words placed in one juxtaposition to other words versus how they are differently interpreted in another.  Beyond that, though, we as individuals have an ability, which has increased through time due to factors I’ll talk about in a moment, to create new context dynamically and independently.  In fact, I’d argue that more than that, sentient beings are distinguished from other beings in their ability to not only create context with respect to new information which they absorb, but via our imaginative facility, we can create context essentially from nothing – and further, we can then endow it with positive meaning to interpret our world (including messages from other people) even though it was created from nothing but an imaginative spark.

That sets up an issue: mathematically, you have two systems of information processing – the context-creative destination, and the context-processing source – that can have no expectation of being aligned well enough to ensure that meaning can be transferred to a destination with the effect desired by the source.  Not only are they in possession of different foundational data (historical information, prior learning, etc.) but they are dynamically creating new ideas or forms or data which may or may not be “real” but will nevertheless impact the destination’s translation of incoming messages and, indeed, will impact the source’s construction of messages.  In the cryptography example, you end up with different codebooks at each end – but more than that, the codebooks themselves are always changing because the sender and receiver are making their own changes to them because they think it would be good / interesting / fun / beautiful to do so.

Okay, so mathematically, we have an unresolveable system; the opposing creating and processing systems cannot be resolved because they are fundamentally always different.  That doesn’t mean no communication takes place, but recognizing this as the appropriate description of systems of communication between two different sentient beings begs the question: how do we manage to be understood in our communications with one another at all?

Luckily, we have a few things working in our favor, or at least they’ve worked to date.  For one, observe that the degree to which we flex our creative potential is normally small relative to what we have in terms of the imaginative capability to achieve in the extreme.  That is, while we have the capacity to come up with radical new ideas and immediately integrate them into our sense of our personal context, in practice, that’s comparatively rare.  More often, we tend to imagine new possibilities which are incremental extensions of what we have experienced ourselves.  If we agree with that description – that our imaginations are infinite but seem to predominate in modest originality rather than radical explosions of the new – then it becomes easy to see that communications, while still fraught with the mutually unintelligible differences between the contexts of the source and the destination, can still be “good enough” if the typical differences in source/destination contexts are relatively small.  When two people who know one another well have a conversation, for example, usually both can make an operative assumption that their understanding of the other context is roughly consistent to what it was the last time they had had a successful conversation, where both felt they had been comprehended and where it had seemed that meaning had been exchanged successfully.

We’re also still evolving, and we’ve only recently emerged from a history in which a number of structural dynamics limited the range of potential divergence in context – at least for most people, almost all of the time.  Those dynamics are worth laying out in a little detail.  First, communication for most of human history has generally taken place within small societies which were largely isolated amongst themselves.  Even if everyone on earth had spoken the same language, they still would have spent effectively all of their time in small groups isolated by geography.  The information available to them – both from the world and from within their own creative capacity – would be limited, and since the groups would have time to reinforce some sort of consensus within the isolated group over time, the range of variability among the contexts of individuals would be small.  Our capacity for creating revolutionary new contextualizations would still have existed, but either they would be incorporated into the common consensus and that group would simply have a new “baseline” context shared by all, or the individual with the radical concept would be rejected – as a witch, as a false prophet, as the insane – and most likely would be killed (early societies are a lot more violent than ours, recent mass shootings aside).

Second, most societies were also broadly pre-literate until comparatively recently, and in such societies the premium placed on memorization and its use in sustaining narrative constructs across time without external storage media needed to be much higher than in more modern societies with broad literacy and, eventually, mass reproduced media.  Small societies which had to employ a huge amount of mental resource on simply remembering in one’s own mind a vast array of social data would have had much narrower personal contextual frameworks – simply because the raw material for ideas was limited, and an individual’s ability to simultaneously remember what they needed to remember pushed out the mental bandwidth for imaginatively constructing new personal context.

In addition, for most of the history of philosophy, there hasn’t been a personalized, internally directed sense of “self” or “mind”; instead, historical constructions of the mind were dominated by a sense of universal values.  Now, they were by no means universal when examined across all of the world, but individual societies or civilizations spent much of their intellectual creativity on discovering external forms.  The freedom we experience in our own minds today, to play with and create new contexts for meaning within ourselves, would have been regarded as folly at best and a sin against the timelessness of meaning at worst: a waste of time when meaning itself should be explored through externalized exploration.  Where it was regarded as a sin – or as a “disorder” of the mind’s rightful search for outward truth – it would have served to quite effectively clamp down on individual imaginative potential.  Even where viewed as a mere folly, however, the likelihood that imaginatively new contexts for meaning would override or dominate the “pretty good” assumptions we normally make when we exchange statements of meaning would be very low – it wouldn’t get in the way of “pretty good” communications, in other words.

Lastly, language itself was much more atomized across groups – there wasn’t a dominance of English or Mandarin or Spanish across huge geographical distances.  Lingua franca emerged, often with highly specialized focuses in commerce or religion or government administration, but that specialization only highlighted the fact that language itself served a strictly localized purpose, either localized in place (different languages on different sides of a range of hills) or in the nature of the communities that used it (traders, sailors, priests).  That meant the expressive possibilities of communication were more limited, because the languages were optimized for small social contexts (even if the individuals who used those languages retained the ability to creatively explore beyond that).  That localization of language served another purpose as well: when different communities did intersect, it was readily apparent to both communicators in an exchange that there was a very high likelihood of failure in the transmission of meaning.  Reflexively, therefore, individuals either narrowed down the range of meaning exchange that would even be attempted, or if there was a need for sophisticated exchange, it would be with high levels of redundancy.  This is directly analogous to what Shannon describes in how data exchange in the presence of high degrees of mechanical interference needs to take place to ensure success: either narrow down the complexity of the message, or broadcast multiple copies of the message to enable the destination to compare the repeated information and sort out what the “true” message is by seeing what over time emerges as being in common among the rebroadcastings, discarding the rest as noise.

Even with these historic conditions in place, however, the generative power of our imaginations to create context from whole cloth would always have prevented perfect meaning exchange from taking place; even in pre-literate or pre-modern times, people are always merely engaging in varying levels of “pretty good” communication.  As we look back in history, we regularly find times where meaning fails to transfer between individuals.  This in essence was the theme of my essay on prophecy last week; when our imaginations create a radically new context that we internalize and make our own, we may for a time be unable to communicate the new meaning as we understand it – or alternatively, others may hear our new contextualization of meaning and be “shocked” by its inconsistency with their given context for translating messages from ourselves into meaning for themselves.  That shock – particularly in settings dominated by contexts which are steeped in tradition – has often ascribed the context of the radically original sender as being somehow divine – the sender in recognizing some revolutionary new context would have somehow received access to the superhuman context of the gods, of the god, or an ontic logos, or our soul, as the case may be in different ways.

So far so good, I think – but as always I’m interested in where other people come at this.  It may feel a bit too mechanistic, and given my starting point from Shannon’s theory of data transmission in the presence of noise, and my being enamoured of the mathematics underlying his theory, that’s understandable.  But as I’ve gone through the exercise of abstracting this into sources and destinations of meaning, and describing those sources and destinations as being bound by contexts which are similar conceptually but incapable of integration in practice, and further, by “humanizing” the source and destination as being sentient creatures with the imaginative capacity not only to learn from new data but to generate entirely new aspects of context without external source – going through this makes sense given what I know about myself when I communicate with other people and what I feel occurs when other people try to communicate with me.

I’m going to apply all of this to the world as we find it today in just a moment, but I want to repeat what I said a little earlier about the philosophical condition of pre-modern people’s contexts for meaning, that is, their lack of a modern concept of an internally coherent self.  Without a dominant concept of the mind or the self as an internally directed, individual space, pre-modern communication was essentially between two individuals who could both be assumed to have an intent of sharing an external context for meaning.  While their particular instantiation of that desired shared context may be quite different between two people – still subject to different context and still subject to the whim of the imagination to expand one’s concept in seemingly random ways relative to one another – the intent of each person served as a further means of reducing the range of noise in meaning exchange.  The architecture of their contexts, if you don’t mind that metaphor, was quite similar, even if the decoration and the surface appearance was varied and sometimes mutually distasteful.  Even those whose contexts which seemed at significant variance with others would have shared an underlying sense that there was still a single model of correct, well-ordered contextualization that existed for all people – and, moreover, which was in reach of understanding to all.  That sense would have been self-correcting in that it would steer the person who was consistently feeling misunderstood or unable to understand back towards “understanding” by acting as a kind of external denial of the possibility of valid but radically different means of applying context to messages they either would create or interpret from others.  The externalization of the “self”, if that is possible as a construct, not only reduced the potential range of misunderstanding, but it also acted to constantly redirect communications back towards mutual coherence.

None of those pre-modern conditions seems to hold today, however.  The rapid destruction of local isolation over the past several centures has had the easily foreseeable effect of eliminating much of the authority of traditional narratives, by exposing people to the truth that there exists an almost infinite variety of outwardly viable narratives which are often contradictory or mutually antagonistic.  I might choose to privelege my local context norms for a variety of reasons – because my local society has more wealth, better technology, or wins more wars, for example, the latter being the typical rationalization – but I can’t ignore the fact that there other successful contextualizations out there.  Exposure to diverse and mutually contradictory external frameworks for truth, moreover, would have eliminated much of the power of a pre-self understanding of external trust to deny individual’s their willingness to accept their own internally derived contextualizations, no matter how far that may have taken them from being understood in their local interactions.  The predominance of just a few languages for online communication, the compression of those online languages (think texting and emojis) to enable those few languages to reach far more than just those who are functionally fluent, and the de-specialization of language and its replacement by specialized jargon within languages all have served to break down barriers of spoken and written communication.

There is a corollary to that, however.  While the genericizing of language now enables more people to engage in meaning exchange with a wider range of people, the former expectation – that the person you would be exchanging meaning had a context for sharing that was pretty close to yours – no longer holds.  Just because someone can attempt to exchange meaning with me doesn’t mean that it succeeds.  This is why it’s so easy to recognize Internet hoaxes coming out of Nigeria: they use the words but the gap in meaning context is wide enough that they are obvious on their surface as false.  They are pretty bad, not pretty good, context equivalents.  That one-two punch, though, where there is an exponential growth in attempts at meaning exchange but a steep drop in the rate of successful meaning exchange vastly increases the amount of energy we need to use to sort the pretty good from the pretty bad.  And if we assume (I’ll go ahead and do so) that each of us also wishes to sort the really good, the transformational good, the stuff that will change our lives good, from the pretty good – which itself takes effort because finding those new contextual meanings that can enhance our lives is hard – then we’re going to run out of energy to keep up the exercise quite quickly, even while we seek to use the freedom offered by these changes to explore new potential contexts for ourselves.

That, really, is the impact of the emergence of the self and the breakdown in belief in an externally derived source of universal context.  Our conception of “truth” today is fluid and personally intermediated by a “self” or a “mind” which is radically and only our own.  I don’t believe this is simply a western phenomenon, either: my readings of eastern tradition is such that, except for radical philosophies which deny that any structure exists, most eastern philosophies also posit an ordered worldview to which our individual contexts are shackled, and that is falling away just as rapidly in China, Korea, Japan, Indonesia, India as they are exposed to the diversity of the modern interconnected world.  The emergence of a self and the release from a belief in an externally structured basis for a single context for understanding gives us incredible individual freedom, and it unleashes the structural constraints on our imaginative potential to create radically personal meaning contexts.  Even if in some totalitarian regimes the “self” is itself constructed by the Party and coerced onto individuals, we observe that the act of coercion in a world of rapid ability to attempt exchanges of meaning tends to have a counterproductive effect.  The coerced “self” is simply the rebel “self” held in check – radically individualized – that’s becoming self-aware through the act of coercion itself.

That imaginative potential is even further unlocked by the fact that our minds no longer have to spend as much energy remembering things any more (even if the work load of processing greater attempts at communication is growing).  It’s been well remarked that the popularization of writing made rote memorization obsolete, and the rise of GPS devices has made an innate sense of direction unnecessary.  Well, our brains haven’t lost any capacity, but they have lost a formerly required use of that capacity, and that freed up capacity can now be deployed in new ways – including towards our imaginative potential.  We have a self, which expands the range over which we can easily roam our imaginative potential – and we have new capacity, which expands the depth and the palette, if you will, of what we may imagine.

So we live in a communication exchange environment where sources and destinations of meaning are infinitely more varied than they were a few hundred years ago, but where the language forms used to communicate are much less varied than they were.  This severely erodes the potential for pretty good levels of meaning exchange by eliminating many of the “obvious” signs which would previously have signalled a need to increase redundancy or reduce meaning complexity.  But at the same time, meaning complexity has increased exponentially, as we deploy our minds with far more creative freedom and far more capacity to enjoy that freedom than in the past.  Greater complexity and the capacity to produce more of it, coupled with a meaning transfer “system” – embracing language and the ability to employ tactical strategies to get towards that pretty good shared understanding – which has put us in a setting where the meaning being exchanged is using less diverse language forms.

Put another way: we’re trying to grasp much more meaning than ever before, both because we live in greater proximity and receive more content than before, and because we’re capable of spontaneously creating that meaning much more easily and freely.  At the same time, we’re trying to use fewer tools to do so, and those tools are suited for simpler meaning exchanges than before even though we need to communicate more complex messages to adapt to and handle the more interconnected world in which we live.

Well, that’s a lovely picture to paint.

I think there are a range of viable ways to extract ourselves from this.  There are a lot of potential solutions but not all are viable: for example, we’re not about to move back to having 4000 languages each spoken by roughly the same number of people, without the kind of ecological or WMD-created disaster that would just send the whole planet back to the Stone Age.  So that’s out.  I also don’t think we’re in a position to reject the notion of the self (or the mind) as being one of the cores of our understanding of the human experience in modernity.  And while there will always be a conservative backlash trying to mount the counterrevolution, I’m also convinced that traditional narrative constructions of shared context are effectively permanently discredited.  We’re also not about to dial back the global experiment on interconnectedness – again, barring some sort of species-wide catastrophe.

So what’s left.  Well, I see three broad classes of response.  One is to recreate the localization of shared meaning in a new, virtual way.  This is at the core of modern identity politics, whether on the left or the right or for social purposes, and it involves paring down your potential communications partners to only those you think may have a similar context for understanding – choosing your audience, in other words.  The idea is that by finding and assembling localizations of people who have an affinity for one another, those groups can create a self-referential zone of pretty good meaning exchange.  Furthermore, they can, either coercively or with incentives, direct the local group’s imaginative activities, which expand the potential for meaning contexts to diverge, in a proscripted way.  This fragments the world as a whole and it seems to increase the likelihood of actual conflict.  I’ve seen this on a personal level but we’ve all seen it on the political level.  Nevertheless, if we accept that meaning can grow infinitely (which I think is broadly true) but also acknowledge that as individuals we have finite capacity to manage our encounters with other contexts, this is an understandable and practically oriented response.

The second approach would be for individuals to learn to restructure their imaginative powers away from creating new contexts and towards imagining others’ contexts.  I’d call this the “mindfulness” option, where we develop disciplines for both internally learning to imagine other points of view as a primary activity, not as a derivative activity for getting “our” point of view more effectively expressed.  You can also think of this as “knowing your audience”, to reflect against choosing one audience in the first potential approach.  By primary activity, I mean spending primary mental effort (or more accurately, effort of our selves) in imagining how our meaning could be reinterpreted or misinterpreted by other meaning contexts, and spending commensurately more effort in reframing how we initiate and participate in meaning exchanges based on this exercise of imagination.  Put differently, we spend more of our energy building a better conception of what other people’s meanings might be, and allocate less towards our internally directed creativity without stifling it too much.  I confess to preferring this approach because it’s not reliant on coercion and doesn’t limit our freedom of imagination per se – it simply redirects how we use that freedom.  And again, it’s our choice to do so, not a mechanical attempt by a local group to rein in or externally direct that freedom.  But it may only be a stopgap; again if meaning can grow infinitely, developing better tools towards grappling with that infinite growth will always be a losing game – or rather, we’ll never stop having to adapt new and better tools, and we may fall behind in the future in a catastrophic way.

The third approach would involve a radical acceptance of our declining ability to manage the infinite growth of meaning contexts, and a similar acceptance of the radical contingency of our relationships with other beings.  We accept that we will not be understood, in other words, and we simultaneously accept that we will not understand others.  I’d call this the “dissolve the audience” response, where collectively we recognize that we’re all simultaneously performers and audience, and we develop a practice of speaking as if we were receiving a message and listening as if we were speaking what is being said.  On its surface this sounds like chaos, and indeed it leaves unanswered questions of how to resolve questions of access to shared or limited resources, how to resolve questions of how we express our human need for “meaning rich” relationships (love, charity, compassion, affection).  Because these other exchanges seem so central to life, an approach to dealing with the radical breakdown in successful communication that doesn’t deal with them seems like a dead end.  This acceptance, however, may offer its own pathway out of that: if we really accept that we will not be understood and that others will not understand us, there should be a degree of sympathy for that shared experience of misunderstanding.  Seeing the unknowable other across the table, or across the screen, then paradoxically enables each person to “know” the other in a much more fundamental way and interpret the otherwise-unintelligble messages through a lens that acknowledges our own frustration at being misinterpreted.  That particle of meaning – of sympathy – would be successfully transferred, not through communications in a directed way, but through an internally understood knowledge of similarity.

Again, that’s compelling because it comes down to personal choice, but unless a critical mass of humanity adopts such an approach, it won’t work.  The early adopters will swim in a sea of alienation and anomie (which arguably is a good description of modernity anyway) in which they sympathize with others but others don’t connect with or understand them.  And their spiritual pain will deter others from trying the same path.  More worryingly, in my experience, those who accept that even pretty good communication is vanishingly difficult to achieve can appear as capricious, or arrogant, or indecisive, or weak to those who are still trying desperately to understand and be understood.  In other words, unless you encounter someone who is also ready to release from the exercise, you run the risk of appearing like a jerk.  And who wants to take lessons from a jerk?

But I don’t see an easy way out – none of these approaches are easy.  Although one could argue that localizing into the small world that I find most recognizable is the simplest response, it still involves broader impacts which are messy and uncomfortable, and to the extent that it leads towards a different form of reining in the human imaginative potential, it seems like a step backwards.  And this is to say nothing of the conflict and hierarchy that is all to easy to see in such localization – hence the breakdown in civic dialogue in modern America or Europe, to say nothing of the lack of dialogue in authoritarian societies.  But relying on personal commitments to either rebalance our exploration of creative potential with more focus on others’ perception of that exercise, on the one hand, or towards a refocused and freely offered sympathy for a shared state of interpretive isolation, on the other, strike me as potentially unachievable – particularly against the pace at which our imaginative powers keep adding to our ability to create meanings as individuated selves.

Personally, today, as I sit typing this, I’m stuck.  It probably means I’m setting up the problem incorrectly.  But I’m also struck that, aside from a brief mention of “meaning-rich” relationships, I still have a problem that my conception of what love is – either between individuals or without object, as an absolute – falls utterly outside of this discussion.  I’ve said before that when I talk about love, it fails – you have to live it and put trust in the lived experience to demonstrate whatever it is you’re feeling.  Like love, there remain aspects to lived meaning for which “communications” in a rational sense are not in scope.  Are those aspects foundational – do we need to accept their existence as first principles?  That feels sort of okay but, like my talking about the divine, it feels like using words to describe a wordless problem.  And if the actual problem is that we can’t resolve lived meaning with contextualized meaning, at least not yet, then any of these potential remedies to the breakdown in communications may just be the critical theory equivalent of replacing the engines on an airplane without wings.

As a father, I’m focusing much more on demonstrating those concepts which are not subject to “exchanges of meaning” instead of worrying about how much my son understands the concepts which I discuss with him.  In that regard I guess I’m demonstrating that I do think of the wordless concepts such as love as more fundamental, on some level than communication.  But I have no doubt my son will be forced to wrestle with this in a far more all-encompassing way than I had to as he grows up.  He will be forced to navigate a world where our imaginations and our physical world have been released to exploring and creating meaning, but where the tools to communicate that meaning to others who also are enjoying that freedom are breaking down.  I don’t know how to help him through that yet.  At least he’s only six – I have a few more years of hugs and voiceless joy before the hard work begins, and I can always lean back on the hugs and voiceless joy when our communication breaks down; my parents still do.  In time maybe – hopefully – he’ll teach me.