Among certain psychologists, there is a belief that the mind is composed of “modules” that have been designed by evolution to account for very specific tasks. This kind of thinking has also been linked to the assertion that the mind is a computer, resulting in the natural rhetorical extension that “modules” are essentially equivalent to the “apps” we have on our phones.Just as your digital device has apps for banking, socializing, navigating, and finding restaurants, your mind has “modules” for tracking resources, socializing, and foraging, or so the argument goes. My own approach to and training in psychology is highly comparative and mechanistic. So, I am sympathetic to the “brain = computer, mind as modules” approach to psychology. I don’t think it is correct, though.
Let’s think this through – and not in an overly academic way. First of all, the brain is not a computer. Both might be machines that deal with inputs and outputs, but many systems deal with inputs and outputs. The solar system, is a collection of matter that handles inputs and outputs in a particular way. A hammer is a system that handles inputs and outputs in a particular way. This does not make the solar system or a hammer computational systems, at least not in any profound way. Similarly, sure, brains and computers share some descriptive features. Both make use of “memory,” both are energy intensive and need a regular replenishment of resources, and both transform information in particular ways. But this does not make brains and computers the same thing. Brains are not computers, even if the brain computes. Ultimately, computers are tools designed by humans for particular tasks. Brains are tools for… well, we’ll get to that.
Secondly, the brain is not composed of modules, even if it is modular. Sure, as we learned in the last post, perception is assembled from “products” that have been created in different areas of the brain. In other words, perception is a distributed process. However, it is a distributed process of shared networks. Just as we discussed with reference to supply chains, different visual “products” do not come from isolated modules. There is no “face” module or “chair” module. Instead, there is a system of shared networks that assemble faces or chairs.
Still, one strength of taking a modular approach to psychology is that it emphasizes the adapted qualities of our psyches. There are indeed deep currents given to our psyches by natural selection. We do seem predisposed to detect cheaters, learn languages, use tools (at least more so than other animals), see and hear a particular range of frequencies, have a sensitized disgust response during the first trimester of pregnancy, and on, and on, and on. Even the most religiously inclined must come to terms with the animal in which each soul resides. Through natural selection, our bodies and minds have been designed to encounter and assemble their worlds in particular ways. This is where the utility of thinking in terms of modules comes in. It allows mind scientists to cleanly carve up their subject matter into the traits and adaptations that allow for research. Every science needs its units, after all.
But let’s not forget that the assembly that is done by our minds is done for a particular reason. Fundamentally the psyche is designed to locate the organism within a problem space. Where am I within this space, and what do I need to do? These are the questions that the psyche faces at every moment of its limited existence. And to answer these questions, the psyche is composed not of modules, but of maps. Perception is essentially a means of creating landmarks, directions, and layers of information on maps. Yes, the supply chains of our perception assemble “percepts,” but on some level, all supply chains, themselves, are mappings of inputs onto outputs. We are not “modules all the way down;” we are maps all the way down. It is not a collection of goods that give our psyches meaning, but directions. Our psyches are composed of, and designed to assemble maps, and in doing so, our psyches search for and achieve meaning.
Our minds are composed of maps, not modules, even if those maps show modularity. Maps locate the organism in problem space, and it it through this that we define meaning.
Neurons as supply chains. This image shows somatosensory neurons of a mouse that have been imbued with a a green flourescent protein. The bright blobs at the bottom are cell bodies, and the rising strands are apical dendritic bundles. In neurons, dendrites are the cell structures that receive inputs from other neurons. “new_20x” by Robert Cudmore is licensed under CC BY-SA 2.0
Since it’s been a while, I’m going to rehash some ideas and then weave in few more concepts: supply chains and maps. We’re still dealing with perceptions and beliefs in all of this.
So, in a recent post, I introduced the idea that emotions and motivations might be more accurately thought of as “visceral beliefs” – a category of perception that originates within the body, and which is built up from basic processes and learned expectations. We call these things “feelings” (I feel angry. I feel hungry. I feel curious. I feel tired…). In each case there is no such thing, really, as say anger. Rather there is a perception originating in our body that we term “anger.” One thing to realize, though, is that even if “you are not feeling it,” this does not mean that your emotional machinery isn’t nonetheless whirring away and influencing your behavior. Not all perception reaches our awareness, just as not all perception is universal. The logic is the same as what is applied to “normal perception” – sight, taste, touch, and so on. So let’s remind ourselves of some of the principles underpinning “normal perception.”
Shipping as neurons, and yes, I’d like you to associate this image with the first image of the post. Paths of shipping in the San Francisco bay. The solid, dark grey areas are land, while the lavender / purple areas are ocean. See this post at Mapbox.Com in order to learn more about how this visualization was created.
Perceptual Machinery
Psychologists long-ago established that perception isn’t a passive transcription of external sensations, but rather involves both a direct construction and an active interpretation. The result of these bottom up and top down processes gives us many of the furnishings and rooms of our psyches. The direct construction is done by the machinery at hand. We have been imbued by evolution with machinery such as rods, cones, chemical receptors, and hair cells that react to a particular range of environmental energies and that turn those inputs into patterns of neural firing. Do all animals, or individuals for that matter, use the same machinery? Of course not. Some birds possess machinery that allows them to “see” the earth’s magnetic fields. Some fish possess machinery that responds to electric fields, and honeybees can see patterns of polarized light that can be used to navigate. As Hamlet would say, “There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy.” The dimensions of our perception are but a sliver of all possible dimensions, and humility (and wonder and a certain arrogance) goes hand-in-hand with doing good psychological science.
Further, even within a common assembly line of perception, e.g., vision, different tools allow organisms to handle different ranges of resources. Our own visual system cannot “handle” electromagnetic spectra from the ultraviolet bands, but honeybees would respond to management’s request to process UV light with “bring it on.” For honeybees ultraviolet inputs are part of their normal experience, while longer wavelengths (i.e., the color “red”) are not (see page 138 of provided link).
To use a crude metaphor, the machinery of perception is like a supply chain that assembles the lived-in environment of the psyche. Depending on the desired product, different supply chains make use of different resources. The supply chain for a rocking chair is quite different from the supply chain for an iPhone. Similarly, the supply chain for a melody will differ from one that produces the smell of cinnamon.
Supply Chains and Perception
Let’s go ahead and push this analogy between supply chains and perception because it will allow us to intuit some general principles of perception.
Here is a map of the supply chain for the Jansport Big Student Backpack. Be sure to click through to the original site, because each of the nodes can be clicked on for further information about the specific factories. Among other things, in this visualization we learn that there is a thread supplier in Malaysia, a yarn manufacturer in India, dying facilities in Thailand, rubber, zipper and yarn facilities in China and Taiwan, and an assembly facility in Indonesia that puts together the final backpack (along with many other products such as shoes, windbreakers, umbrellas, etc.) before shipping it to a primary distribution center in California. This distribution center then introduces the backpack into consumer awareness by placing it on the racks of a store or in the pages of a catalog. The consumer “sees” the final backpack, not the undergirding supply chain that makes the backpack possible.
Our own perception works according to similar principles. Dedicated areas of the brain specialize in “producing” vertical lines, horizontal lines, colors, visual movement… , and yet other areas receive these inputs and assemble them into larger products. At some point these perceptual products enter our awareness, but there is a considerable amount of assembly and processing that has already happened prior to this point. Below, for example, is a crude diagram showing the “supply chain” for visual information in Homo sapiens. Interrupt that supply chain at different locations and particular visual goods will no longer be available.
A map representing the supply chain of vision in humans. It is an adaption of Figure 4 from Behnke’s Hiearchical Neural Networks for Image Interpretation. Notice that visual supply chains classically separate into two crude pathways: one that assembles WHAT something is, and one that assembles WHERE something is. For our purposes, prosopagnosia could be thought of as involving disruptions to the WHAT supply chain. Blindsight, on the other hand, involves damage to the V1 area of the supply chain…a critical choke point for conscious perception of visual information. Nonetheless, in many blindsight patients visual factories are still hard at work in the SC, pulvinar, and from the inter laminar portions of the LGN. Additions to Behnke’s diagram are based on a number of readings, but anyone interested could start with Cowey 2010 and Fulton’s 21’st Century Paradigm Describing the Neural System (p. 129). The latter is kind of eccentric, but for this post you only need to worry about the visual circuitry diagram on page 129. Basically, we now have a better understanding of the pathways from the lateral geniculate nucleus (LGN) and pulvinar nucleus that by-pass the V1 area of visual processing. Also, this diagram is still extremely crude. Not shown, for example, are the many feedback pathways that are known to exist among the different cortical areas.
Prosopagnosia, for example, is a phenomenon in which individuals lose the ability to “see” faces (click here to read an article about potential treatments). prósōpon = face, a = not, and gnosis = knowledge. Prosopagnosia usually results from damage or congenital alterations to a brain region known as the fusiform gyrus. Individuals with prosopagnosia might be able to describe the components of a face just fine: the nose is bulbous, the eyes are brown, the hair is curly and black. However, despite possessing the components of a face, they simply don’t possess the ability assemble those components into a face.
The condition known as blindsight also illustrates how disruptions in the visual “supply chain” are not an all-or-nothing thing. Blindsight is characterized by a functional blindness in which individuals nonetheless possess abilities that are dependent on vision. (If you have access to online articles, here is one that provides a nice overview and here is another.) An individual with blindsight might not be able to see a baseball, but could potentially catch one tossed their way. They might not consciously know that an object is present, but nonetheless could correctly guess as to the direction of its movement. In general the reason for the blindness of blindsight involves damage to area V1 of the visual cortex. This is an area at the back of the brain that receives the bulk of input from our retinas. However, as the figure above shows, several inputs from our retinas are sent in parallel to other brain regions (the superior colliculus and the lateral geniculate). It is thought that these inputs allow for the “sight” of blindsight. Conceptually, though, we can think of this with reference to our supply chain analogy. If we lose our “distribution center” or our “assembly factory” certain types of perception might never reach our awareness. Nonetheless, this does not stop the thread, zipper, or fabric dying factories from producing their goods – goods that can be used in other supply chains, as needed.
Figure from “To See, But Not to See.” A nice representation that uses the supply chain analogy to describe how disruptions along one pathway still allow for “deliveries” to take place along other pathways. It just so happens that only one of these pathways involves conscious awareness (the one indicated in blue).
Issues in supply chains don’t only have to do with the loss of resources such as we “see” in individuals with blindsight or prosopagnosia. For example, what would happen if a supply chain began to include inputs of colored fabric to an assembly factory that normally only received black fabric? We might end up with rainbow colored backpacks! Synthesia resembles just such an outcome. This syndrome refers to a phenomenon in which some individuals experience the stimulation of multiple sensory pathways when only a single pathway is objectively stimulated. Numbers and sounds might be “seen” to have particular colors; observing another individual’s hand being touched, might produce a sensation of touch in the observer. It should be noted that synesthesia is not an imagined or arbitrarily conjured sensation. Rather, the synesthete simply and consistently perceives a mixed sensory experience relative to what others experience.
I don’t want to push the analogy between the machinery of perception and supply chains too far. After all, it’s just an analogy. Computers aren’t the same things as brains and supply chains are not the same thing as perception. However, there’s one more point I’d like to squeeze out of the supply chain / perception analogy, and that’s the concern with balancing resilience and efficiency.
A central issue faced by the management of supply chains involves balancing efficiency and resilience. For example, the coronavirus outbreak has made news for disrupting supply chains through China, and many green technologies currently face vulnerabilities in their supply chains that might affect their ability to scale up. In order to handle these issues capable managers need constantly to grow and prune their supply chains in ways that minimize costs while maximizing resilience. Extra supplies need to be maintained, backup factories need to be kept at the ready, and of course all of these contingency plans impose costs. If the world were perfectly predictable, these costs would be unnecessary. But the world is not predictable. Storms, viruses, political unrest, new technologies and competitors all impose costs. Such is life.
Supply chain or brains? In this case it’s neural networks, but conceptually the figure originates from work that seeks to emulate the synaptic pruning observed in brains. Taken from Singh (2019). Pruning deep neural networks.
Perhaps not surprisingly, the machinery of perception deals with similar demands for resilience and efficiency, and the way that our brains manage these demands shapes the environs of our individual psyches. For example, the development of our brains is characterized by an overproduction of connections, followed by a pruning of these connections. This “blooming and pruning” occurs in different areas of the brain at different times in our lives. Below, for example, is a figure showing the approximate developmental time course for several brain regions along with their hypothesized functions. What it shows is that our visual and auditory “supply chains” are crudely mapped out with a hodgepodge of connections that are then tightened up over the first three years of life. Areas involved in speech production, on the other hand, have a later “blooming” that is followed up by a pruning that occurs until approximately age 7. [Remember: these are averages and say nothing about what is happening in the brain for any single individual.]
Figure 1 from Thompson and Nelson (2001). Developmental science and the media: early brain development. American Psychologist, 56, 5 – 15. The curves estimate the number of synapses in different brain regions at different times in an individual’s life. The central point is that waves of “blooming and pruning” are a regular part of brain development across (at least) the first 20-odd years of life.
The pruning of pathways in our brains is crudely guided by a “use it, or lose it” maxim. Connections that are repeatedly stimulated are strengthened and maintained, while those that are rarely stimulated are deleted. The result is that our perceptual supply chains are made more efficient at delivering the products that we apparently need. If you do not need to distinguish between “l” and “r” sounds, then your brain will eliminate the wasted circuitry that produces the perception of these two distinct sounds. Simultaneously we become worse / better at associating particular lip movements with particular sounds. Similarly, as I brought up in another blog post, if you do not need to distinguish between faces that have particular racial characteristics, then your brain will eliminate the “wasted” circuitry that produces the perception of these differences. In extreme cases, experimenters have even been able to create environments in which animals lose the ability to perceive horizontal or vertical lines. Again, use it or lose it. If the world does not need you to “see” vertical lines, then your brain is not going to waste the supply lines necessary to create a perception of vertical lines. In psychology this phenomemon is termed perceptual narrowing and it is the result of synaptic pruning, but conceptually perceptual narrowing and synaptic pruning are no different from what supply chain managers do every day.
Summing up the Supply Chain / Perception Analogy
All right, the main principles to draw from our analogy between supply chains and perception are:
Just as supply chains construct products by integrating and orchestrating a network or factories, the machinery of perception constructs our psychological world from many distributed areas in the brain.
Just as a zipper or yarn factory can help in the assembly of many different goods, perceptual “factories” can provide outputs that are used for many different types of percepts (e.g., faces, objects, animals, etc.).
Just as much of a supply chain is invisible to the consumer, much of the machinery of perception is invisible to our consciousness.
Just as supply chains are molded by concerns for efficiency and resilience, our brains mold the machinery of perception by balancing physiological costs and environmental demands.
