Why Did Humans Evolve Pattern Recognition Abilities?

Evolution necessitates economical and reliable brain mechanisms in some environmental contexts. These mechanisms emerge as a response to patterns in the environment or enable us to refine our ability to spot them. Pattern recognition skills sit at the helm of our basic cognitive architecture. Why? How? Now what?

A common problem during hunting is to estimate how many predators there are – based on cues like animal sounds, footprints, etc. Say a pack of 4 hunters is trying to isolate a prey for food. The hunters can only survive if they have the physical capability to defend themselves and successfully kill or escape. If they do not have the ability, they will die. The ones who survive must have had the ability to recognize patterns, identify cues, and take action.

Say 10 wolves are lurking around the corner. Four hunters can hardly fight them. The hunters can then decide to stay back and wait. This scenario plays out only if there is a specific sensitivity to magnitude estimation – can the hunters correctly estimate how many predators there are? If the hunters can’t distinguish between 1 and 10 wolves, they might escape or kill one wolf, but against 10, they will die. If they can differentiate the pattern of sounds & environmental cues which estimate the magnitude of predators, they can make decisions that aid survival. The ability to recognize the differences in patterns created by 10 wolves and 1 wolf is useful to survival. As a result, our senses adapted alongside our cognitive ability to make sense of sensory signals.

The survivors survived such scenarios when they could differentiate contextual cues. This is a primary type of pattern recognition – estimating magnitude in proportions. 

Now, 1 wolf is easy to count, 2 is easy, 3 is easy, 50 isn’t. But when 50 wolves are present, it doesn’t matter if you are off by 10 wolves. In essence, there are many wolves, exactly how many is not essential. You are going to lose. 

Think about walking through a mountainous field. But now, imagine there are mountain tops in every direction. Can recognizing patterns help you decide where you want to go? A particular 3-4 mountain-top arrangement can be easy to identify, but a 20 mountain-top pattern might not be. When it is 20, the overall structure with a few landmarks can suffice in estimating the direction of the destination. Instantly recognizing 3 to 4 features is easy because we developed a specialized cognitive template to recognize small numbers without counting.

This specialized tendency, along with many other contexts like fruit spotting, leaf spotting, honed in a general trend to subitize. Subitizing is the Human ability to count up to 4 items at a time without calculating or breaking down into sub-parts. Precision in identifying small quantities is important but its importance goes down as the quantities increase. Humans can get the ball-park for large numbers well enough to estimate its influence. If there are about 200 bees – run away. If there are 5, run away slower.

These moments created a network of general and specific pattern recognition abilities. Let us look at more. But before we do that, how was the opportunity to have these abstract cognitive skills created in the first place?

The leading theory is that the brain developed many of its higher functions because it had space to grow, and these higher functions evolved alongside many other templates of thinking – enumeration, vocalizing, tool-refining, etc.

When humans learned how to cook meat with fire about 1 to 2 million years ago, their massive faces with enormous jaw muscles were no longer providing any advantage. Cooking helped to predigest food and make it easier to bite. It even helped us sanitize raw meat and reduce bacteria-based illnesses.

These jaw muscles took up neural space & skeletal space. According to genetic evidence, generations later, the weak-jawed ones survived because that space was taken up by the expansion of our cerebral cortex – the folded wrinkly part of the brain which is involved in decision making, remembering, speech, risk analysis, innovation, inferring causality, questioning, etc. This expansion marked the beginning of a more “cognitive” era.

Inside a skull, there is limited space; to fit in additional neural connections, these neurons begin folding. And now, thanks to those folds called (gyri and sulci), we have art and technology. While other animals have gyri and sulci too, it appears that it is only one of the conditions needed for cognitive expansion. So not all creatures with gyri and sulci develop human-level cognition.

A unique example of pattern recognition is facial recognition. Most humans could identify human bodies from an assortment of other animal bodies, but when tribes formed, in-group & out-group differentiation became important. That’s when it was essential to know members’ faces. Just about 200 neurons from 100 billion neurons can efficiently map the entire human face based on a few focal points – a few near the eyes, jaw, chin, eyebrows, etc. (study on monkey brains)

Pattern recognition in sound was also important to our ancestors. Our voices modulate along with emotions and this association has a social advantage. Without sharing any more information, listening to human voices can predict others’ emotional states. Emotional states predict motivation, behavior, ability, need for care, danger, etc. Giving automatic special attention to screaming & crying is also a pattern recognition-based survival advantage – to provide aid & care when one hears it. 

Ever imagine why certain melodies sound the saddest on a violin than any other instrument? The timbre of violins (sound characteristics) is congruent with human cry vocalizations.

Many different brain regions are implicated in pattern recognition and we don’t clearly know how the process is organized. Some evidence suggests that the orbitofrontal cortex acts as a representational system for the features in a pattern. The hippocampus might be the hub for generating a memory of elements in a pattern. The occipital extra-striate cortex is involved in subitizing and other primitive attention-independent pattern recognition. Other regions like the superior parietal cortex and the insula are involved in integrating attention, sensory-motor movement, emotions, and various cognitive processes. It functions as a hub for connecting action, emotions, and cognition.

How have these evolutionary tendencies affected us today?

This question brings us to cognitive biases. Most of the pattern recognition skills humans developed had a context, and that context has changed today. But, the brain retained those tendencies. A lot of “jumping to conclusions” is based on these primitive patterns. You may have heard of the confirmation bias. Here is a likely evolutionary explanation for why it exists.

Confirmation bias – we identify and attend to information which agrees with our belief. If you believe a plant is dangerous, it’s good to jump to a conclusion to avoid it and be wrong than try it out and die. True positives and negatives in identification are great. False negatives can be deadly, but false positives can keep you safe. 

Here are some more cognitive biases (many are derivates of the confirmation bias).

How does memory and attention factor in? As far as we understand the brain, patterns turn into abstract memories which may be inherited or developed over time. But they aren’t isolated memories; they are linked with our attentional system (the ability to focus and attend to information from our senses). Patterns are natural attractors, and they attract our attention. In a way, our attention is fine-tuned to attend to patterns because there is a memory of that pattern across the brain. This memory primes our attention to identify it. Think of it this way – your attentional system is always on and readily available to identify significant patterns. 

Ever seen a cat jump at a long thread and get excited? It’s because of an embedded snake pattern that was once useful for a cat’s survival.

That is why it is easy to spot Jesus in Coffee mugs and clouds. It is also why humans tend to project their ideas and thoughts on ambiguous structures (Zodiac signs in the sky). Why that turned into astrology is a different story.

Along with attention, humans developed a signal-to-noise ratio filtering system, which made identifying patterns easier. Humans can filter out useless information and focus on useful information. Scribble a few lines, and you can choose to see image patterns in it by ignoring some scribbles. Humans tend to assign meaning too and complete incomplete pictures. An incomplete picture can be imagined to be complete by imposing a pattern on it. Gestalt psychology principles highlight this (the whole is greater than the sum of its parts). Focus on the next image, do you think the designs are similar to complete shapes?

How about this image? Spot the mobile phone.

Throughout evolutionary histories, correlations were valued because they showed consistency. If some things happened regularly (deadly wasps after rain) and you responded in a particular way (hide), it was best to continue doing so. Correlated events are hotspots for pattern recognition advantages because humans process this information as patterns. 

A recent hypothesis is that spatial memory is linked to the sense of smell. One of the reasons why this hypothesis has gained some evidence is that smell helped us navigate the world with respect to safety – food, waterfalls, flowers, dead animals, etc. all have identifiable smells that informed us about the safety of a particular route. Brain regions involved in smell and spatial memory appear to be linked as well, and damage to those can impair both abilities. This finding shows how deeply patterns can be co-dependent.

Patterns need not be exact but they tend to capture an approximation. We began to impose patterns, even if they didn’t exist. The famous golden ratio hype^[1] showcases this. We can assign a pattern to just about anything because it helps us conceptualize information.

You may be more familiar with reflexes – built-in behavioral responses that occur as a reaction to some distinct event. These are patterns of behavior that emerge as a response to a sensation. Remember we talked about our attention being ready to “attend” to certain types of informational patterns? Reflexes go beyond that. Attention is not mandatory – reflexes operate in the background, usually without involving the newer regions of the brain. Sensory nerve cells get stimulated which activate inherited responses called reflex arcs, bypassing conscious interactions in the brain. Here, pattern recognition occurs at an unconscious level because that makes it reliable, less variable, and consistent.

Let’s take an example. When you touch something extremely hot, your skin recognizes a pattern of information/stimulation via pain receptors (nociception) and triggers the withdrawal reflex – you withdraw your body instantly. Now, this pattern is barely a pattern. It’s more of a characteristic. As these characteristics get more and more complex, a pattern may emerge and that pattern is embedded as a heritable memory.

Assigning patterns itself is a cognitive skill that survived evolutionary hurdles. The utility of pattern recognition was so high that it survived and permeated in virtually every context in life.

In short, pattern recognition was a cognitive advantage which helped in survival and growth because these thinking patterns were:

• efficient
• super-fast
• automatic
• reliable 
• usually right in a context
• helped avoid lethal risks 

These necessary pattern recognition skills were embedded in the brain, and they demanded lesser cognitive and biological resources. They helped us to avoid risks in many contexts and survive. 

Where does this take us ultimately? Patterns made noisy information meaningful for the brain. In fact, the first layer of meaning itself may be the existence of a pattern that guides behavior.

References:

Chang, L. J., Yarkoni, T., Khaw, M. W., & Sanfey, A. G. (2012). Decoding the Role of the Insula in Human Cognition: Functional Parcellation and Large-Scale Reverse Inference. Cerebral Cortex, 23(3), 739–749. doi:10.1093/cercor/bhs065 

Duchaine, B., Cosmides, L., & Tooby, J. (2001). Evolutionary psychology and the brain. Current Opinion in Neurobiology, 11(2), 225–230. doi:10.1016/s0959-4388(00)00201-4

Konovalov, A., & Krajbich, I. (2018). Neurocomputational Dynamics of Sequence Learning. Neuron, 98(6), 1282–1293.e4. doi:10.1016/j.neuron.2018.05.013   

Logan, G. D., & Zbrodoff, N. J. (2003). Subitizing and similarity: Toward a pattern-matching theory of enumeration. Psychonomic Bulletin & Review, 10(3), 676–682. doi:10.3758/bf03196531

Sathian, K., Simon, T. J., Peterson, S., Patel, G. A., Hoffman, J. M., & Grafton, S. T. (1999). Neural Evidence Linking Visual Object Enumeration and Attention. Journal of Cognitive Neuroscience, 11(1), 36–51. doi:10.1162/089892999563238 

Stedman, H. H., Kozyak, B. W., Nelson, A., Thesier, D. M., Su, L. T., Low, D. W., … Mitchell, M. A. (2004). Myosin gene mutation correlates with anatomical changes in the human lineage. Nature, 428(6981), 415–418. doi:10.1038/nature02358 

Dahmani, L., Patel, R. M., Yang, Y., Chakravarty, M. M., Fellows, L. K., & Bohbot, V. D. (2018). An intrinsic association between olfactory identification and spatial memory in humans. Nature Communications, 9(1). doi:10.1038/s41467-018-06569-4 

P.S. At the foot of the right leg. On the bluish patch.

Sources