programming, rude squirrels, and meditations on learning

Programming still makes me feel like a dunce.

I stare at the display of my computer. Its bluish light reflects off my corneas, but I am not seeing the bluish light. The words on the screen denote a script that I have been trying to write for a couple of hours. It should help me organize a data structure with different forms of a survey I am using, along with the questions, answer choices, and correct answers. It also should create unique IDs for the questions and answers. But the code isn’t quite done. My fingers brush circles on the table. Its surface is subtly bumpy. I’m not sure of what the next step is. I draw paths on the surface between the circles according to my planned trip through code-space.

Actually, one more push of work might be enough.

The letters on the screen denote the items and the groups to which they belong. The lines of text are staggered in order to represent this nesting of groups. That is, question represents a question on the developed survey, and question has several properties. question_text denotes the text of the question, question_difficulty denotes the difficulty of the question, choices is a list denoting possible answers, and answer denotes the correct answer. Each question is in a list called question_list, which is in turn a property called form_questions of a larger group called form, which also has a form_number and grade_level. Each form is in a list called all_forms. Each question of course inherits a form number and grade level by virtue of being in the form that it is in.

To create an ID number for each question, we first need to make an ID property for the question which will later be filled in with a value. To generate the right value, we approach this list of forms. As we enter the outer layer of abstraction, there are different properties, one of which is a list. Items in the list cycle past until the correct item is reached and selected. Then, we can enter to a deeper level of abstraction. Namely, that item has more properties, including lists, each with their own items which whizz past until the appropriate item is selected and lights up. This item has yet another set of unique properties. The correct properties are now selected. In the distance through the translucent layers, a faint glow flickers on. The connection appears stable.

A muted, repetitive sound. A sound? The innermost layer of abstraction deflates and recedes into a small speck as my submerged head zooms outwards away from it, passing through the next layer of abstraction. My head passes through the next layer and then the next, back out past the inner workings baked into each level like a synthetic glutinous cake, the pressure around my head reducing with each of the levels, which squelch past until my head finally resurfaces with a final jerk of my neck. The muffled beeps are coming from my alarm. I shake my head, flicking the remnants of colorful abstraction goop out of my ears, which return to full functionality and flinch at the alarm. These are some terrible beeps. I turn off my alarm. In the absence of terrible beeps, my ears pick up the whirring of the computer’s fan hard at work. I blink. I’m at my desk with its barely bumpy surface. The words denote my progress on my script.

I should save my progress. A few more clicks of the keyboard and my script is saved, computer off, and with a bit more stumbling I am at the door, slipping on shoes. It is time for my scheduled walk. I don’t think I’ve forgotten anything. I think I’ve got everything I need.

I forgot: it’s pleasant today. My eyes take a moment to adjust to the sunlight when I step outside and start my usual walking route. The birds greet me as I cross the street toward their tree. The tree beckons the breeze with its branches, and the breeze nuzzles my cheek.

Sure, this was hard work for me because I am still a novice programmer. Honestly, part of me twists itself up knowing that I didn’t have to spend so much time on this. I might have finished more quickly if I did it all manually instead of programming at all. But I want to get good at this. I know that a more experienced programmer could probably finish this script I spent hours on today within around thirty minutes. They have a rich system of concepts and rules crystalized in their minds, ingrained in their fingertips. Those are the same concepts and rules that I still struggle with. Some days, it’s impossible to believe that if I just practice, I could have that system too. I know that the only way to get that expert knowledge is to leap into the depths, gain experience, and then organize it in my mind, reorganize it, and reorganize it again. And yet, as a novice, I often don’t have the knowledge to know if I’m making any progress. Learning deeply is a huge, uncertain investment, and I’ll admit I find it intimidating.

By the same token, the strenuous and unstable nature of learning also makes it totally transformative. When I emerge from doing this type of learning, I feel my life change fundamentally. I can use my new knowledge to reformulate problems that were previously incomprehensible and find solutions to them. I learn a new set of values and develop new habits in accordance with them. It even transforms my basic perceptions. Things have attributes that I didn’t see before, and I become attuned to this additional dimension in my environment. I don’t mean to make this sound mystical. Maybe if I describe the phases of learning I cycle through, it’ll make more sense. Consider the following analogy.

PHASE ONE: DISARRAY AND HYPERFOCUS

Descent

In the first phase of learning deeply, I adjust my goggles, take one last breath, and dive. When I open my eyes, I am in a strange realm, surrounded by strange shapes, textures, and colors. Everything seems at once in organic, dynamic unity and in completely incomprehensible formlessness. This duality becomes increasingly evident as I dive deeper. Everything is new. I cannot tell where one object ends and another begins. I don’t know what I’m looking at, I don’t know where to look, I don’t know how to look. My eyes narrow, focused, to try to find similarities between objects. At some point, they have glazed over in their overwhelm.

Then, I start to recognize objects that I saw before. There are duplicates, quite a few actually. I think I’ve seen most of the objects this place has at this point. Most of them are familiar or at least similar to the objects I’ve already seen, just a different color or size. Now that I have a sense of the variability in color, size, shape, and other attributes, I can imagine what other kinds of objects could be here. Also, it seems like certain types of objects are consistently close to each other down here. I wonder if that’s a coincidence. When I search around for one of the objects, the connected object is always in close proximity. These two objects must be connected in some way. This might be a pattern. The two objects are close together when the arm of the first one is moving in the direction of the other object.

But, when I simulate this potential relation between the two objects, incorporating the variability in object arms and the pattern with arm direction, something feels missing. What makes the object’s arm point in the direction of the other object? Is the second object drawn in some way to the first object’s arm? Or does the first object’s arm detect the second object, and then direct itself towards the second object? Or is there some other outside factor that influences the behavior of these objects and their arms? Wow, there’s a huge cluster of these objects over there! What is happening?

What I’m trying to get at is that, in the first phase of learning deeply, I dive headfirst into a new area and absorb as much information as I can. Then I can group objects based on whether they are similar or different in a particular way. This way of being similar or different between objects can be summarized as an attribute, such as shape or color. An attribute “flattens”, or reduces dimensionality of, the problem space. I can also start to see relations between objects based on their attributes and mentally simulate these relations to generate possible explanations. I can look around and collect information to confirm or disconfirm these patterns. 

 

Patterns allow me to abstract from my experience, that is, to flatten my experience into concepts and rules that I can carry in my pocket back to the surface. When my pockets are full enough, my expedition is complete, and I can start to make my way back to the surface.

Phase One, the descent phase, requires my mental resources to hyperfocus on the objects in the environment: observing, detecting similarities and differences, hypothesizing relations, and testing beliefs. In order for hyperfocus to be possible, another cognitive task needs to occur in the background: suspension of doubt and indirectly related knowledge. My mind is temporarily suppressing my other indirectly related knowledge which could distract me from understanding the information in front of me or induce undue doubt in the patterns that I perceive. This suspension helps me to maximize my focus on the information in front of me, to be open-minded and to give it the best chance possible. As I begin floating back to the surface, the cognitive load of learning reduces, and the mental force suspending my other knowledge and doubt begins to loosen. Memories of previous expeditions flit back into my consciousness.

PHASE TWO: INTEGRATING PREVIOUS EXPEDITIONS

Ascent

A few miles away is the spot where I did another expedition months ago. That time, I was surrounded by strange objects with a completely different shape from the ones from this dive. They were huge. Those objects were covered in little tuber-shaped appendages which were changing color. I thought my eyes were malfunctioning. Actually, I now recall noticing that, as in this trip, the objects sometimes came in pairs. In particular, the appendages of paired objects were a milky blue color, unlike the color-shifting appendages of the other objects. Although the objects seemed different, there are similarities between the previous and current expeditions. Both of these environments are home to objects that seem to be trying to find a partner. And they’re not that far away from each other, just a few miles. Maybe there was something in the environment, some mechanism that could account for this. I wonder why color was relevant in the last expedition, while the arm direction was relevant this time.

In this phase, the new patterns or rules that I learn in one context start to trigger memories of other patterns in other contexts. I can start to integrate the new patterns with the other patterns I learned in the past by finding similarities or differences in the pattern itself or between the contexts in which the patterns emerged. Similarities between patterns or contexts help me to create a general rule that seems to apply in both contexts. This is valuable because I can try to generalize the rule, to start to imagine what kinds of environments I might observe the pattern in, even if I haven’t observed that environment before. On the other hand, noting differences is also useful, because I can question what mechanisms could bring about these differences. This information helps me to identify important distinctions that I might observe in other contexts, which would inform my theory.

This remembering process is integral to consolidating and reorganizing specific knowledge into more general structures that could be used in multiple contexts. Having these more general concept structures, including general rules and relevant distinctions, allows me to think of new questions to ask, which can motivate subsequent expeditions. They can also forge new connections that spark creativity. Successful integration of new information deepens my understanding of phenomena I learned about in the past. The infrastructure between concepts in my mind transforms and sees a boost in connectivity. To the extent that my mental life depends on the infrastructure of concepts, my mental life thereby transforms.

In other words, learning deeply is transformative for me in part because of the nature of the phases in the learning process: hyperfocused sensemaking and comparison-based integration with previous experience. These two phases differ in two key ways. Phase 1 has a narrow scope in my mind and requires utmost focus to make sense of the seeming chaos of information. On the other hand, Phase 2 has a larger scope encompassing the new learning as well as memory of previous learning that might be relevant. Phase 2 also requires a relaxing of focus on internal workings of concepts in order to freely notice new similarities and differences. They each have their benefits, but their benefit together outweighs the sum of the individual benefits.

Inability to move from Phase 1 to Phase 2 results in learning of information that is not integrated with my prior experiences and knowledge. The new information doesn’t feel relevant to anything else that I have learned. It does not spark novel questions, it does not boost connectivity, and it does not induce creativity. Spending time only in Phase 1 is not only unproductive, but also disconcerting. Once the novelty of learning new things ebbs away, I get a nagging sense that the new information is not useful or meaningful to me. It’s important that I spend some time allowing my mind to wander off after Phase 1, for it will come back with memories that are, in some way, relevant to my new knowledge. Making connections to these memories will make this learning meaningful with respect to other concepts and phenomena.

On the other hand, spending time only in Phase 2 and no time in Phase 1 stunts my learning. If all I do is scour for relations between information I already know, then I will eventually reach a point at which I can’t think of any more, and any new relation I can think of will be relatively unimportant and stale. The issue is that I will be out of new material from which to abstract concepts and relations. This can only be by going on new expeditions. Phase 1 expeditions require intense focus, but the return is fertile experience, ripe with potential to enliven my concepts and theories in Phase 2.

This learning process requires that I operate my cognitive faculties at full capacity. I start to lose track of what time it is or how long I have been working. I lose touch even with my bodily sensation and surroundings. The bluish tinge of my computer display, the texture of my desk top, the pain growing in my unwisely angled neck. All is suspended, inhibited so that mental resources can be fully allocated to learning.

Even my sense of self disappears. Although the reality is that I am trying to understand and learn a new thing, I lose awareness of my self as an agent that is trying to understand. There are not two separate entities, a mysterious phenomenon and me (the observer trying to understand it)—rather, there is a continuity between the thing and me, and in light of this continuity, in that moment, me, my body, my passage through time, my agenthood are all unimportant. All there is, is the thing at hand. I “become” it, embody it. As I learn its inclinations, its ambitions, its observations, its limitations, these become, in a sense, my inclinations, my ambitions, my observations, and my limitations. I am lastingly transformed.

Hang on. Did you hear that? What on earth is that sound? The breeze ruffles the leaves. There it is again, that squawking sound! It came from up there above me in the tree by that squirrel! The squirrel looks me dead in the face, opens its mouth, and squawks again. “Excuse me?” I am not one to be condescended to, and I could tell from his eyes that he was sizing me up. He squawks again. The nerve. Does he not realize that before he shouted at me, I had been thinking? Or is it that he doesn’t care? He doesn’t care that I’ve got things to do, thoughts to think, work to complete? He is a rude squirrel. I glare at him. The rude squirrel glares back.

My eyes narrow at the wind. The rude squirrel seems oblivious to it (as well as all of the other important things in the world).

I can see my reflection in his unrepentent eyeballs. My eye is twitching and now I can’t help but admit that my face looks ridiculous.

I’m a moron. Of course this squirrel doesn’t care about interrupting my thoughts. This squirrel doesn’t care about me. To the squirrel, I’m a strange animal, too large an animal to be a friend and too slow to be a foe. With one final squawk, the squirrel turns its puffy tail on me and scurries up the tree.

I’ve even had my earphones in the whole time without playing any music. When I’m in my thoughts like this, it’s easy to put everything else on autopilot, which I guess means talking back to rodents. The wind is picking up. The sky is preparing to change from baby blue to velvet purple. It’s probably time to make my way home. It’ll be dinner time soon.

I don’t even remember what I was–oh right. Yes, OK. So, when I am learning deeply, especially about new topics, I seem to lose myself. While this can be an, exhilarating, transformative experience, it can have negative consequences.

When I embody the phenomenon, connection to myself disappears, and so does my control over myself. This lack of personal awareness and control allows me to plunge into a new territory with an open mind and to absorb as much as I can. But, this also makes me do silly things (like seeing malice in the eyes of innocent squirrels) and worse. If I spend days on learning deeply and nothing else, conceptual goop perpetually in my ears, I lose track of the fact that I’m lacking nutrients, sleep, or exercise. I forget to do leisure reading, socialize, and make and consume art, the things that maintain and delight my being. This doesn’t work because in the long term, my being is all that I am. Sure, I may be transformed in some way by my new knowledge, but the other more foundational facets of my being are being neglected. The weakened foundation renders the other facets of my life, including my thoughts, simulations, and concepts, lifeless. My first-person experience loses continuity and coherence. Unless I can remember who I am and why I am learning in the first place, it all feels meaningless.

PHASE 3: REMEMBERING THE SELF

To shore

In this phase, I come to my senses.

I’ve floated back up, almost to the surface, and realize that work is over for the day. My head emerges from the water and blink the salt water out of my eyes. My arms and legs feel shaky and heavy. Diving is serious business and puts a lot of strain on these muscles. They need time to recuperate. More than anything else though, I’m hungry. Actually, I’m ravenous. Time to go home.

As soon as I get home, I heat up some soup and rice and sit at my table. My arms flop down at my sides. My muscles release and slump. I’m glad that dinner was pre-prepared, and so are my floppy muscles. Eating at the end of a long day always makes me feel better. Today was definitely a long day. I reflect on what I learned. I am feeling content about my work today and proud of my progress. There are times when I am also ashamed of the clumsier moments of my work and try not to think too hard about it. But most days, I feel proud of what I’ve accomplished. A few more expeditions and maybe I’ll have enough information to write up a report about the trends I’ve noticed in my dives. Tomorrow, I can start planning the next expedition, and then maybe by the end of the next month I can start writing. For now, my ears are begging to listen to music. I would also like to call my grandmother. Maybe I can read a short story before bed.

The final phase of learning deeply involves mindful healing and careful introspection. This includes reflecting on the learning process, including successes and mishaps, and how they relate to my ambitions, as well as attending to my more basic physical, emotional, and social needs. In Phase 1, my attention is hyperfocused on a narrow scope, a specific phenomenon. In Phase 2, my attention becomes broader to include a larger scope, such as prior knowledge and experience. In Phase 3, my attention is distributed or diffused across an even broader scope, my entire first-person experience in that moment.

In Phase 3, my mental resources are directed from the outside world back to me. This allows me to become aware of how my body and mind feel. This information is important because it helps me to identify what I need to do next. In the short term, I may need to eat, sleep, call a friend or my family, write a scathing email (and not send it), sit in solitude, reassure myself of my abilities, or devour some celebratory potato chips. I also reflect on decisions or actions that I may need to make in the long term. For example, I might realize that I would like to change my topic of research. Alternatively, I may find that I am captivated by my topic of research and would like to pursue it by taking on new research methods or developing my own. I may find that I want to spend less time on work and more time contributing to social justice causes.

Phase 3 gives me space to perform personal maintenance, but it also helps me to establish or see the coherence in my life’s events. The shift from Phase 1 to Phase 2 helps me to see the coherence or conceptual continuity between the new thing I learned and the old things I learned in the past. That is, these deep diving expeditions in learning aren’t just weird arbitrary tasks; they relate in important ways to other things I have learned and done. In the same way, the shift from these two phases to Phase 3 brings to light the fact that I am not just learning or working hard for no reason, but because learning and working help me to pursue my curiosities and acheive my goals. That is, learning has a context: me, the learner. Contextualizing the act of learning in me specifically creates a continuous, coherent story from the events in the life of an agent with goals moving through time.

If I spend too much time in Phases 1 and 2, I lose my self, I become tired, and my life starts to feel disjointed and incoherent. I don’t remember why I cared to do the things I do in the first place, but I still do them. The seeming lack of reason behind my behavior makes wonder if I am crazy. My physical or emotional needs start to accumulate, and I become perpetually tired. On the other hand, if I spend all of my time in Phase 3, I become very well attuned to any shifts in states in myself, but I cannot transform myself with experiences or learning and cannot gain insight from my knowledge structures. I fall out of the practice of learning deeply or making connections, and I cower at the prospect of learning, analyzing, or deliberating. The best way to facilitate my learning seems to be to cycle naturally through these phases.

Again, consider the acuity or depth of attention on one hand and breadth or size of scope of attention on the other. In Phase 1, my focus is deep and the scope is small. This is a tradeoff needed because attention is a scarce resource. Spending an entire day in Phase 1 would look like a long, skinny cylinder. In Phase 2, focus is less deep, but there is greater breadth, as connections between different concepts are made. Spending an entire day in Phase 2 looks like a shorter, more stout cylinder. In Phase 3, focus is completely diffused, not hyperfocused or fixated, across my personal experience, a large breadth. Spending time here looks like a large flat cylinder, like a coin. The volume represents the amount of learning achieved. If I try to do all three, but without taking breaks in between, my day looks like three cylinders stacked on top of each other. The volume of this is the sum of the three component parts.

Being able to move between these three phases routinely establishes continuity between them, filling in the gaps between cylinders, and this looks more like a cone. This cone has the virtue of capturing differently sized scopes, while also capturing different depths of focus. By virtue of being a more continuous shape, it also has a larger volume.

Of course, in practice, learning doesn’t always happen from Phase 1 to 2 to 3. It can oscillate between them. My goal here is mostly to define three important phases of learning, and to emphasize the need for cohesion in moving between them.

What I am trying to say is that cone-shaped learning allows me to learn deeply, widely, and meaningfully. Cone-shaped learning is more meaningful and more sustainable way to learn and grow and be, because I can move between phases with cohesive transitions. I can’t always seem to get through all of these phases in a day—it usually takes a week or more—but even just nudging myself between them using alarms has helped.

It’s definitely helped today. I think I’ve learned a lot: I finished a new script and practiced moving through levels of abstraction in code. I think I can be proud of my work. I shuffle across the grass to get the mud off my shoes. No mail today. My stomach moans. I guess I’m starving. When I open the door, my nostrils are not met by the smell of freshly cooked rice. Damn it to hell. I forgot to start the rice cooker before my walk. I wish that squirrel yelled at me sooner. I am a dunce indeed.