Marcus became interested in the human mind as a teenager. In High School, after creating a program which translated Latin into English, he came to the conclusion that one cannot build programs within machines that understand language without understanding how people can understand language. This led to his progressive interest in cognitive psychology. He attended Hampshire College where he designed his own Major: Cognitive Science working on human reasoning. He continued on to graduate school at Massachusetts Institute of Technology where his advisor was the experimental psychologist Steven Pinker.
Marcus's research and theories focus on the intersection between biology and psychology. How do the brain and mind relate when it comes to understanding language? Marcus takes an innatism stance on this debate and through his psychological evidence has given many answers to open questions such as, "If there is something built in at birth, how does it get there?" He challenged connectionist theories which posit that the mind is only made up of neurons. Marcus argues that neurons can be put together to build circuits in order to do things such as process rules or process structured representations.
Marcus’ early work focused on why children produce overregularizations, such as breaked and goed, as a test case for the nature of mental rules.
In his doctoral dissertation, Marcus studied how children acquired the past tense of English verbs. He looked at 11,500 utterances by children to see when their past-tense forms were right, when they were wrong, and what the circumstances were. Although children knew how to use the past tense default (adding –ed to the end of a verb) they were unable to do so with verbs they did not know.
In 1999, he discovered that 7-month infants have the capacity to acquire abstract rules, such as the ABB structure in sentences such as la ta ta and wo fe fe.
In his first book, The Algebraic Mind: Integrating Connectionism and Cognitive Science (MIT Press, 2001), Marcus challenged the idea that the mind might consist of largely undifferentiated neural networks. He argued that understanding the mind would require integrating connectionism with classical ideas about symbol-manipulation.
In his second book, published in 2004, The Birth of the Mind: How a Tiny Number of Genes Creates the Complexities of Human Thought, Marcus goes into a more detailed explanation of the genetic support systems of human thought. He explains how a small number of genes account for the intricate human brain, common false impressions of genes, and the problems they may cause for the future of genetic engineering.
In 2005, Marcus was editor of The Norton Psychology Reader, including selections by cognitive scientists on modern science of the human mind.