Language in the Brain #2

Diversion into categorisation: it’s all about forgetting


 [Is this your first time reading this blog? If so, you may want to check the previous post out :) ]

Another long train + another + an amazing espresso from THE coffee place in London (Prufrock Coffee -these guys have won World #1 Barista and can seriously pull a mean shot!) later, here is the next instalment of the “Language in the brain” series. In the last chapter we left our working definition of language in the following state:

“Language is a social and conventional system for the externalisation and internalisation of our thoughts.”

But, if you remember, we finished on an open-ended series of questions (purposefully left open so as to leave the reader wanting a bit more –oh, clever device!). Those questions came up because we mentioned the fact that language seems to entertain a very close link with our mental states and that simple utterances of words seem to be able to capture so much more information than we thought. Anyway, we’re not talking about this today because to fully grasp the added value that language offers to cognition, we will take a few steps back and consider a key and central process of human cognition: categorisation and what better than an interesting anecdote to get us started?

Writer Luis Borges, in his 1962 novel, depicts a fictional character, Funes, who after a bad fall from his horse develops a neurological disorder giving him infinite rote memory. As a consequence, Funes would never be able to forget anything. Every instant of his life would be stored forever and he could browse through the collection of memories very much like you would browse for that specific moment in your favourite DVD. Pretty handy right? Well, one would think… Indeed, Funes was puzzled by the fact that people around him would call a dog that he had seen at a particular time, in a particular place, in a particular orientation, under a particular light, and so on, by the same name as they called it a few seconds after because its orientation had changed for example. His infinite memory was therefore a great handicap because, for him, every instant was stored as a separate entry in his memory and all situations appeared as unconnected and genuinely different. Funes’ ability to abstract over experiences and remember the essential invariants of entities was gone. This simple process that we take for granted, this amazingly useful ability to forget in order to make sense of our experience is in fact no other than the process of categorisation, which we should agree, seems to be essential to human cognition across the board.

Psychologist Steven Harnad contends that categorisation is so central to making sense of perceptual and internal experiences, so central to any kind of learning, so central to thinking that, in fact, “cognition IS categorisation”. This idea, although it may seem quite strong and radical at first pass, is not so surprising when we look into how our brains have evolved circuitry that is almost specifically designed to facilitate categoriation processes.

The story of specialisation starts as early and at as basic a level as our visual system in the Inferior Temporal Cortex (ITC) which features neurones that have complex shape specialisation confirming that as early as basic perception the notion of categorisation is present where subsets of the visual system treats stimuli differently (Li, Mayhew , & Kourtzi, 2009; Muhammad, Wallis, & Miller, 2006). Areas in the Prefrontal Cortex have also been found to treat different aspects of stimuli such as boundaries and conjunctions between those as well as abstract rule-based categorical distinctions: categorisation. The Medial Temporal Lobe seems to be particularly involved in storing regularities and exceptions to rules: categoriation (Love, Medin, & Gureckis, 2004). Finally, another crucial feature of the human brain is its differences in plasticity rates where while subcortical regions have fast plasticity leading to quick changes in synaptic weights (quick learning and storage of information) higher cortical regions, which are slower in plasticity, are not tied to specific events and require repeated exposure to change synaptic weights and effectively will only be affected by a sort of average of stimuli: once again, categorisation (think of it as superimposing pictures with some transparency where only the common and recurrent information presented in the images will emerge, see figure 1)

Figure1. Illustration of the generalisation derived from repeated exposures to different perceptual stimuli.  Adapted from stimuli in Thierry et al. 2007. 

[Note that of course what really happens in the brain is not as trivial as this but I thought it was a good illustration.]

In fact, plasticity is so crucial to categorisation that imbalances in plasticity (due to disruption of dopamine receptors) often lead subcortical regions to be stronger than normal and effectively overwhelm higher regions by details, making generalisation impossible. Such disruptions have recently been correlated with psychiatric and neurophysiological such as autism (Dölen et al., 2007).

I hope that from that (very) brief overview of the biological bases for categorisation you will be able to gauge how much of a prominent and essential feature it is for cognition in that it is instanced in the very basic organisation of the human brain. And naturally, at this point, since this blog series is about language, you might ask what this article has to do with language. In fact, you might even argue that the processes and structures I have reviewed so far are not all specific to humans. To this I would answer with the following question: if those processes are not human-exclusive how can human cognition be so much more advanced than that of other mammals? And, because I’m keen to create suspense, all I can add now is that this will be the topic of the next chapter. In that chapter, I will propose that humans’ advanced cognition does not rely on just better and more efficient brain organisation, but that something else (you have guess it: language) is highly likely to be where a significant part of this advantage resides. À bientôt!

Further reading:

Seger, C. A., & Miller, E. K. (2010). Category Learning in the Brain. Annual review of neuroscience, 33(1), 203–219. doi:10.1146/annurev.neuro. 051508.135546


Dölen, G., Osterweil, E., Rao, B. S. S., Smith, G. B., Auerbach, B. D., Chattarji, S., & Bear, M. F. (2007). Correction of Fragile X Syndrome in Mice. Neuron, 56(6), 955–962. doi:10.1016/j.neuron.2007.12.001

Li, S., Mayhew, S. D., & Kourtzi, Z. (2009). Learning Shapes the Representation of Behavioral Choice in the Human Brain. Neuron, 62(3), 441–452. doi: 10.1016/j.neuron.2009.03.016

Love, B. C., Medin, D. L., & Gureckis, T. M. (2004). SUSTAIN: A Network Model of Category Learning. Psychological Review, 111(2), 309–332. doi: 10.1037/0033-295X.111.2.309

Muhammad, R., Wallis, J. D., & Miller, E. K. (2006). A comparison of abstract rules in the prefrontal cortex, premotor cortex, inferior temporal cortex, and striatum. Journal of cognitive neuroscience, 18(6), 974–989. doi: 10.1162/jocn.2006.18.6.974

Thierry, G., Martin, C. D., Downing, P., & Pegna, A. J. (2007). Controlling for interstimulus perceptual variance abolishes N170 face selectivity. Nature Neuroscience10(4), 505–511. doi:10.1038/nn1864


Leave a Reply