The Extension of Neural Complexity

In the last chapter of his book Networks of the Brain, Olaf Sporns extends his neural processing and, thus, cognition beyond the brain and to the body and the body's environment. This is the feature of neurophysiology that finally destroys all my old ideas about cognition, thought, and knowledge, for no longer can I think of thoughts as belonging only to the brain. Thoughts and emotions – all forms of cognition – flash through the brain, through the body, into the environment, and then back through the body and into the brain. I have only to think of some of the lively and spirited conversations that I have had over the years to see how my thoughts at any given time were not my brain's alone, not even mine alone, but the reiterative, feedback process of patterns flashing through the conversational space from my brain to my colleague's brain and back to me and back to them, over and over. Sporns, of course, says it more scientifically precise:

By acting on the environment, the brain generates perturbations that lead to new inputs and transitions between network states. Environmental interactions thus further expand the available repertoire of functional brain networks. … The body forms a dynamic interface between brain and environment, enabling neural activity to generate actions that in turn lead to new sensory inputs. As a result of this interaction, patterns of functional connectivity in the brain are shaped not only by internal dynamics and processing but also by sensorimotor activity that occurs as a result of brain-body-environment interactions [which] can be conceptualized as an extension of functional connectivity beyond the boundaries of the physical nervous system. (306)

Sporns follows the argument of Andy Clark to say that

the minds of highly evolved cognitive agents extend into their environments and include tools, symbols, and other artifacts that serve as external substrates for representing, structuring, and performing mental operations. If this view of cognition as extending into body and world is correct, then cognition is not "brain bound" but depends on a web of interactions involving both neural and nonneural elements. The networks of the brain fundamentally build on this extended web that binds together perception and action and that grounds internal neural states in the external physical world. (309)

Those who are familiar with Stephen Downes' thoughts on this issue (for example, here) will quickly recognize his ideas about the extension of knowledge through a social network, so that anyone person's – say, Susan's – knowledge of the French capital Paris is a network of flashes across Susan's brain, body, and interaction within the general, historical discussion about Paris as well as with the actual city of Paris. For Susan, then, cognition of Paris is the interplay of patterns in her head, in her body, in her conversations with others (mediated by voice, text, image, networks, and other media) and with Paris itself. Indeed, the more sophisticated Susan's Paris network becomes, the richer is her repertoire of ways to think Paris. At any one time, Susan will likely never use the entire network of meaning available to her, but because she has such an extensive, rich network, then she can think significantly about Paris in almost any situation for any reason.

I have a couple of quick observations to make about this view of knowledge as a kind of cognition. First, we can only have personal knowledge. By that I mean that Susan must always view Paris from the center of her meaning network. With lots of training and hard mental work, she can perhaps learn to look at Paris from other points of view than her own, but she can never not think of Paris from her own point of view (I think that's the correct combination of negatives. Count'em). Even if she changes her mind about Paris, she is simply knowing Paris from a different center, but still her own.

Second, knowledge can never be merely personal. Yes, this contradicts my first observation, but there it is. Susan's knowledge of Paris always extends throughout her ecosystem to include shared language, shared social groups, shared experiences, and so forth. Susan must form her knowledge from the center, but she must also form it in dialog with others who are likewise working from their own centers. Any attempt by Susan to look at Paris from another's center is a sometimes useful exercise in fiction. It's a God's view that Susan can sustain for only a short time. Any attempt by Susan to look at Paris only from her own center is a fatal entrapment in fiction. Knowledge depends on what Morin terms the dialogic principle: the constant interaction of any entity from its own center with its environment and the other entities in that environment interacting from their own centers. Knowledge is that zone of tension between loss of self in its own center and loss of self in the centers of others. Susan interacts with her world – sometimes skillfully, sometimes not so skillfully – and that's what makes Susan who she is. Education is the attempt to help Susan interact more skillfully.

Complexity and Cognition

You might think that complex systems are complicated, but they often aren't.

That may be a bit too cutesy, but it does make a nice distinction between complexity and complication in network systems. Modern jet fighter planes and computer circuit boards are complicated structures – they are composed of millions of parts arranged in intricate ways for a myriad of purposes – but they are not complex. Why? Because they don't change, and if they do change, then that change usually breaks them. They are rather rigid structures, with regular, predictable, and reliable interactions among their parts. After all, you don't want a jet fighter that suddenly decides to start behaving differently in a dog fight.

On the other hand, complex structures such as the human body change constantly, acquiring new cells, functions, and capabilities and discarding old ones. They are dynamic, and not just in the sense of moving parts. They are dynamic in the ways the parts within the structure interact with each other and in the ways all those structural parts interact with the ecosystem that encloses them. And the trigger for this dynamism is sometimes quite simple. In Networks of the Brain, Sporns paraphrases Herbert Simon to say:

First … most complex systems can be decomposed into components and interactions possibly on several hierarchical levels. Second, complexity is a mixture of order and disorder, or regularity and randomness, which together account for the nontrivial, nonrepeating nature of complex structures and their diverse dynamics. (279)

Brains, then, owe their neural complexity to "the union or coexistence of segregation and integration expressed in the multiscale dynamics of brain networks" (278), to the mix and tension of "some degree of randomness and disorganized behavior with some degree of order and regularity" (281,282), and to "rich and dynamic contextual influences" (286). This dynamic complexity in the brain is what gives rise to the emergent property of consciousness, or as Sporns says it, "Consciousness emerges from complex brain networks as the outcome of a special kind of neural dynamics" (298).

I see, then, two elements that generate neural complexity and, thus, consciousness:

1. nodes and clusters of nodes – from single neurons to social networks and natural ecosystems – that are able to form meaningful patterns within any given scale and across all scales (segregation and integration of functions)
2. a fluid tension between regularity and randomness, order and chaos, as patterns form, fade, and reform across the web of nodes as nodes form their own patterns and then harmonize those patterns with the other patterns forming, fading, and reforming elsewhere in the neural network

I need a better picture, so I'll call again on the image of the brain as two musical groups: a left hemisphere orchestra and a right hemisphere jam band. Imagine the New York Philharmonic meets The Allman Brothers Band on the same circular, floating stage: the Philharmonic stage left, the Brothers stage right. The musicians can hear both each other and the speakers that circle the stage, filtering and focusing the sound from two omnidirectional microphones pointed out toward the world. (The musicians can also see, feel, smell, and taste, but let's not overcomplicate this metaphor. Sound will suffice, I think). Finally, they have microphones on stage through which they can play, or not, their sounds to the outside world.

Both bands are mature. They know their chops, their instruments, and each other. They know how to make music on their instruments and how to blend their individual music into the music being made by the other instruments on the stage AND to the music coming in over the speakers from the outside world. When they are all rested and focused, then they can make wonderful sounds that harmonize internally with the other sounds on the stage and externally with the sounds coming over the speakers from outside. When they are not rested or they've had too much to drink, then they make silly, discordant sounds, sometimes truly awful sounds.

Because they are mature musicians, they are dedicated to learning more about their instruments, each other, and their music, so much of the time they are focused on their internal, on-stage noodling, trying this new combination of instruments, this new musical motif or riff, or practicing and honing old motifs and riffs to have ready at hand when they need them. They have a huge repertoire of different sounds that they can call upon at an instance, and they know which among them can make which sounds. None of them can make all sounds, and some of them can make only a few sounds, but they all know how to group and regroup themselves as needed. Sometimes they group as strings, which will pull together the violins and guitars, sometimes as low register instruments, which pulls together the tubas and bass guitars.  The point is that they have a rich repertoire of established sounds, and they are constantly working to add to that repertoire.

But they are also keenly aware of the sounds coming from outside, and they will respond to sounds they hear. They can faithfully reproduce and harmonize with sounds that they know and other bands with which they've played before, creating a pleasing musical interlude, melodies and movements arcing back and forth between the different bands both on a single stage and across the different stages. 

This is where it gets fun. If you are lead guitarist Duane Allman (an individual neuron in the right hemisphere of the brain), then you are listening to your bandmates Dicky Betts, Gregg Allman, Butch Trucks, Jai Johnny Johnson, and Berry Oakley AND to the New York Philharmonic just across the stage with tonight's guest cellist Yo Yo Ma AND to the sounds coming from the other orchestra/jam band made up of the Boston Pops and the Grateful Dead. You are listening for a place for you to fit in. You at last hear a space for you and you make your sound. It's a particularly pleasing, clever riff, so Yo Yo Ma echoes it. You echo back. It's picked up by Phil Lesh of the Dead, reworked slightly, and comes back to you again. You restate it, then rework it again, expanding it by a few bars. The woodwinds in both orchestras join in, and the musical pattern soars. Everybody's happy. Everybody understands the same thing. The band has created a pattern of sound that you, Duane Allman, could not have produced alone but that could not have been produced without you.

Or perhaps you make an awkward sound, something that just doesn't work in the current flow. The musical pattern becomes chaotic for a moment until the other musicians ignore you. The music rights itself, and the bands move on as Brother Gregg leans over and whispers to you, "We're playing in G, dude."

Why did Gregg do this for you? Because – and this is the most important point – there is no conductor, no central processing unit, no boss. The musicians (the individual neurons) are all on their own, seeking a way to integrate their individually produced contributions into the whole. They are each guided by their own, unique abilities to produce unique sounds and by a shared interest in harmonizing, synchronizing, and otherwise fitting their sounds into all the other sounds to create a pleasing, workable whole.

In resourceful, well-tuned bands (brains), each musician finds a way to fit into the whole, most of the time playing a supporting, complementary role, sometimes taking the lead, but always looking to add her own unique sound to the group and its music. In damaged or deranged bands, the musicians are stuck playing the same tune over and over, or they cannot integrate with each other so that no coherent music emerges from their individual sounds.

If I understand Sporns, this is how cognition takes place: emerging, dynamic patterns of firings of individual nodes that group, fall apart, regroup in clusters across the left and right hemispheres of an individual brain AND across different brains, mediated by our actions and symbol systems.

So what does this view of cognition mean? Well, for me as an English teacher, it means that if I want to understand fully the meaning of Shelley's poem Ozymandias then I must be mindful of the complex interactions within Shelley's mind, the complex interactions of the symbol system he used to compose the poem (English language, poetry, sonnet, etc.), the complex interactions of Shelley with his ecosystem (natural, social, intellectual, etc.), the complex interactions of the poem with its ecosystem (production printing, distribution, consumption, etc.), the complex interactions of Shelley's readers with the poem and with each other, the complex interactions of all those interactions with me, and the complex interactions within my own mind.

This effectively describes an approximately infinite number of dynamic interactions that I must master in order to understand completely one fourteen-line poem, and it's why we can write about one poem for two hundred years and still not exhaust it. Basically, what it shows is that we can form a richer understanding of Ozymandias, but we cannot form a complete understanding. This is a crisis for students confronted with a regime of objective tests. In the face of this crisis, students do the only sensible thing: they either demand to know the correct answer, or they walk away. We force them to choose either to play the one correct tune over and over or to degenerate into chaos. Let us hope they are able to choose wisely.

CCK11: Knowledge and Context

If you are like me, then you don't have too much trouble imagining cognition as a network phenomenon: all thoughts, visions, dreams, calculations are based on the spidery webs of firing neurons flashing in unique patterns like lightning through our brains. I can imagine, then, some peculiar and unique flash of lightning in my brain each time I think of, for instance, Connectivism. Each time that flash of lightning fires, my brain creates the concept Connectivism, and re-traces those routes along those particular neurons, across the various regions of my brain, traversing both hemispheres, so that I can think Connectivism. In this scenario, a particular flash of brain lightning equals a particular concept, and I can reinforce that concept by flashing it again and again in different contexts until I firmly etch the pathways into the circuitry of my brain. Nice image.

But wrong. If Sporns and his fellow researchers are correct, then this network of firing neurons is too regular and static. It appears that the brain is much more complex than that, and it is possible that any given idea such as Connectivism is not fixed to any specific network of firing neurons. Rather, the brain may use different neurons over different times to create the same pattern of meaning, depending on what else is already happening in the brain. It seems, then, that as we cannot find a specific chunk of knowledge in our brains, we also cannot find a specific network pattern of knowledge in our brains. The brain is far too dynamic for that. As Sporns points out, specific brain functions are not tied to specific brain regions, nor are specific regions tied solely to specific functions:

Different complex functions are accomplished by transient assemblies of network elements in varying conditions of input or task set. In other words, different processing demands and task domains are associated with the dynamic reconfiguration of functional or effective brain networks. The same set of network elements can participate in multiple cognitive functions by rapid reconfigurations of network links or functional connections. (182,183)

Sporns concludes that "functions do not reside in individual brain regions but are accomplished by network interactions that rapidly reconfigure, resulting in dynamic changes of neural context" (183). This suggests to me that any given bit of cognition depends very much on the interaction between the bit of cognition and the neural context within which it immediately finds itself seeking expression. Sporns says that "the functional contribution of a brain region is more clearly defined by the neural context within which it is embedded [and] this neural context is reconfigured as stimulus and task conditions vary, and it is ultimately constrained by the underlying structural network" (183,184).

So what is the takeaway lesson here? For me, it is this: the brain is a most complex orchestra of two parts—a right hemisphere jazz band and a left hemisphere classical orchestra, and each time it expresses the concept Connectivism, it chooses different instruments and different musical arrangements, depending on what the rest of the orchestra, and the conductor, and the audience are all doing. Sometimes, the concept Connectivism may find expression in my brain from the classical side, expressed mostly with woodwinds and a single flute. At other times, the concept may find expression from the jazz side, expressed through a wailing saxophone, a drum kit, and an electric fretless bass, with some contribution from the classical brass section. Either way, or in some other way, I can still recognize the musical motif: the idea Connectivism, but it is not the same network pattern expressed invariably each time. Each time it is expressed, Connectivism is something slightly different, even in my own mind.

This reminds me, of course, of Edgar Morin's view of complex thought, that "the intelligibility of the system has to be found, not only in the system itself, but also in its relationship with the environment, and that this relationship is not a simple dependence: it is constitutive of the system" (11). Sporns' research into neural networks reveals to me that complexity functions in the brain's production of a single thought which depends on the interaction of any defined unit with its enclosing unit and with all the other units that it encloses and that enclose it. A concept depends on the interaction of a given neuron with its brain region, that brain region with the other regions, all those regions with the brain, the brain with the rest of the body, the body with its enclosing groups, and so on.

CCK11: The Orchestra of Mind

In Chapter 9 of Sporns' book Networks and the Brain, I think we reach the heart of the issue for the discussion about Connectivism. In this chapter, Sporns is tackling the issue of cognition, or neural activity in all its various forms: learning, thinking, feeling, daydreaming, dreaming, etc., and he makes the bald, bold statement that "cognition is a network phenomenon" (181). This is the basis of Connectivism. It is certainly consistent with Stephen Downes 2008 statemtent in his Innovate article Connectivism & Connective Knowledge that "the term connectivism describes a form of knowledge and a pedagogy based on the idea that knowledge is distributed across a network of connections and that learning consists of the ability to construct and traverse those networks." Sporns lays a strong, well researched, authoritative foundation for the discussion of Connectivism, and I think it's helpful to consider three traits of this network concept of cognition as Sporns lists them at the end of Chapter 9:

1. "Cognition has an anatomical substrate" (205).
2. "Integration involves dynamic coordination (synchrony, coherence, linear and nonlinear coupling) as well as convergence" (205).
3. "Stimuli and cognitive tasks act as perturbations of existing network dynamics" (206).

Anatomical Substrate

I find it easy for my own New Age imagination to posit some cosmic Mind emerging from the idea of cognition as network, but Sporns assiduously avoids any hint of the New Age, keeping his conversation firmly grounded in the physical anatomy of the brain and its observable and verifiable behaviors. As he says, "All cognitive processes occur within anatomical networks, and the topology of these networks imposes powerful constraints on cognitive architectures" (205).

These neural networks have "small-world attributes" that are similar to other networks that are perhaps more familiar to us: social networks, the Internet, Wikipedia, or gene networks (Small-world network). Small-world networks are characterized by clustering, which groups nodes about a more well-connected node, thus facilitating quick connectivity to most any other node in the large-scale network. This is similar to the way Google works on the Net: it is a well-connected node that reduces the hops between us as individual nodes and most any other node on the Net. It may be the way an orchestra works with clusters of violins, bassoons, and drums, with a lead violinist about whom the other violins cluster and who connects those violins to the rest of the orchestra.

This small-world architecture means that cognitive processes are both segregated and integrated. An act of knowing relies on segregated clusters of neurons firing in different regions of the brain (different instruments sounding in different regions of the orchestra), and these segregated firings are then integrated (apparently within milliseconds) into a coherent thought, much like all the sounds of the orchestra. Which brings us to Sporns' second trait of cognition as network phenomenon.

Dynamic Coordination and Convergence

The brain apparently has a couple of mechanisms for integrating segregated patterns of neuronal firings into a coherent whole pattern. First, the brain can synchronize neuronal firings, or perhaps the better way to say this is that the neurons can synchronize themselves, much as a jazz band can find its way back to a  chord progression, time signature, or tune to synchronize the various flights of solo fantasy. Then, various firings can converge in a particular firing which takes the inputs and feedforwards them as a single input.

I do not understand neurophysiology well enough yet (perhaps never) to understand how synchrony and convergence work in the brain, but I understand the import: the brain is an incredibly rich orchestra of instruments which can sound or not, in unison or not, according to an internal rhythm, pitch, timbre, tone, and volume, and the brain's clusters of instruments—say, the woodwinds—can converge to create a single tone amongst the tones from the other regions of the orchestra. Every single instrument is vital to the orchestra, but no single instrument makes the orchestra. Every instrument must maintain its own integrity (the trumpet must not try to become a cello) just as surely as it must find its place in the orchestral network.

Stimuli and Cognitive Tasks as Perturbations

I think Sporns is using perturbation in its scientific sense as a variation or deviance in a system caused by some outside effect, and this is perhaps the most amazing trait of cognition as a network phenomenon. We in education typically think of external inputs (classroom lectures, for instance) as the most important aspect of learning, yet if Sporns is correct, then external inputs are best seen as an interruption of or as a more-or-less complimentary addition to the internal system. As I noted in a previous post, any external input must earn its place in the orchestra. An external input hardly ever totally supplants the music the orchestra of the brain is already playing, except in the case of trauma or other profound experiences. Of course, anyone who has ever tried to get the attention of twenty-five kindergarteners (or of twenty-five kindergarten teachers, for that matter) knows how difficult it is to keep them focused on your external input, whatever it is. This, it seems, is the natural state of things. Education is not the systematic, mechanical input of data into the blank data banks of students' minds; rather, education is an attempt to join and to modulate the tunes already playing in our students' minds. Surely these two processes are radically different and require radically different pedagogies.