Abstract
One of the many strengths of humans is the ability to create and understand intricate languages of reading and writing. Scientists have long debated how humans’ brains have developed these reading and writing specific skills in such a short time. Neuroscientists’ recent study from Massachusetts Institute of Technology, however, has unveiled that rather than our brains evolving to perform linguistic functions, the inferior temporal cortex (IT cortex) has been “recycled”. In functional magnetic resonance imaging (fMRI) research, a region within the IT cortex known as the visual word form area (VWFA) lightens when the brain recognizes orthographic stimuli, in this case, words. Additionally, researchers used fMRI to find that areas of the IT cortex meant for face and object recognition lit up distinguishing words after learning to read. This shows that the human mind is adaptable and can repurpose itself for different tasks, and in this case, for reading.
Experiment
To test this theory, Old-World monkeys, such as baboons and rhesus macaques, which diverged from humans about 25 million years ago, were studied. Despite not being perfect models for the human mind, these primates have enough similarities to be comparable. Research has shown that we share similar functions and structures in the ventral visual pathway which is a section of the brain utilized for object recognition. In prior research from 2012, baboons were able to differentiate real English words and nonsensical combinations of letters or non-words with enough training. This proves that word recognition is something that does not require years of evolution and a complex understanding of linguistics. However, researchers still pondered the neural workings behind this skill.
This curiosity led to a new test, where scientists used microelectrode arrays to record the neural activity in untrained macaque monkeys while they examined both words and non-words. These arrays were placed in the monkeys’ IT cortex as well as a part of their visual cortex called V4, which connects to the IT cortex. Then, they put the data into a linear classifier, a computer model designed to identify whether or not words triggered neural activity in the monkeys. Dr. Rishi Rajalingham, the lead of this study, explained that this process is very effective and easy, because there is no training necessary for the monkeys.
Data
In this diagram, Model A presents an example of text the macaque monkeys would view, a visual of the IT cortex, and V4 lighting up. On the right, the figure shows which orthographic tests affect which specific neurons using data from the microelectrode arrays.
Model B displays three different sets for the monkeys. The first contains samples of words and random assortments of letters, and the second tests the variation in the size and spacing of the words. The third set displays one letter but each card has the letter in different locations.
Model C presents the location of the microelectrode arrays in 4 example monkeys: N, S, B, and M. To the right are 8 different graphs representing example IT sites, each with some of the monkeys’ neuronal response to a set of 5, consisting of both words and a random string of letters. The colors of the data show the intensity of the response with blue being the lowest and red being the highest. In addition, there is shading around some sections of the data that represents the general margin of error. Over the 300 millisecond time frame, many of the words sparked higher responses in the monkeys compared to the jumbled letters despite having no prior training or experience with orthographic stimuli.
The model displayed that the data corresponding to the IT cortex was about 70% accurate, with both performance and errors made by these monkeys similar to the baboon study in 2012. Meanwhile, the visual cortex was notably less accurate which reveals that the skilled areas of the IT cortex for object recognition are specifically capable of being remodeled if needed for reading. These observations reveal that the IT cortex in untrained monkeys is sufficient enough to complete simple orthographic tasks such as word recognition. In the future, researchers plan on studying both trained monkeys and literate humans to compare results. With adequate training, our brains are capable of learning to read without highly evolved brains and by “recycling” our IT cortex.
References
The inferior temporal cortex is a potential cortical precursor of orthographic processing in untrained monkeys, Nature Communications, August 2020
https://www.nature.com/articles/s41467-020-17714-3
“Parts of the human brain have been ‘recycled’ for reading, indicates study”, News Medical Life Sciences, August 2020
“To read, humans ‘recycled’ a brain region meant for recognizing objects”, United Press International, August 2020
“Key brain region was ‘recycled’ as humans developed the ability to read.”, MIT News, August 2020
http://news.mit.edu/2020/brain-recycled-ability-read-0804
Orthographic Processing in Baboons (Papio papio), Science, April 2012
https://science.sciencemag.org/content/336/6078/245.abstract
Kyle Phong, Youth Medical Journal 2020
Youth Medical Journal 