Elucidating the Brain’s Language Circuitry with Dr. Michael Ullman

Yair Ben-Dor
May 16
6 min read

Updated: 24 hours ago

Georgetown neuroscientist Dr. Michael T. Ullman's Declarative/Procedural (DP) model posits language relies on two core brain memory systems. Developed at MIT, his widely cited theory links our mental lexicon (words) to declarative memory (facts/events) and mental grammar (rules) to procedural memory (skills). This framework, derived from studying neurological patients, offers key insights into language processing, disorders, and learning.

Architect of the Declarative/Procedural Model

Michael T. Ullman, Ph.D., is a professor in the Department of Neuroscience at Georgetown University and director of the Brain and Language Lab, which researches the neurocognitive bases of language and memory. While earning his doctorate in Brain and Cognitive Sciences at the Massachusetts Institute of Technology (MIT) in the early 1990s, Dr. Ullman proposed the declarative/procedural (DP) model, a theoretical framework that provides key insights into the neural bases of language learning and use. His work is widely cited and has applications for the study of the neurocognition of language, memory, and other cognitive domains.

Dr. Ullman’s Path: The Road to Neuroscience

Dr. Ullman’s interest in science stems from his childhood experiences. “When I was a kid I was always interested in nature and science. My best friend’s father, in San Francisco, was an early geneticist, and I thought, ‘Oh, genetics would be cool,’” he said. “When I went to Harvard, there was no genetics major, but there was a biochem major — the closest I could get — so I majored in that at first.” But after beginning his studies, he said, “I just didn’t enjoy being in a lab with Petri dishes; it wasn’t my thing.” After his sophomore year, Ullman took four years off of school to help start up a computer company and then travel internationally. He then returned to Harvard. However, he switched his major to computer science, with a focus on cognitive science and language. He initially hoped the field would prove useful in launching another company, this time in machine translation.

After delving deeper into the science, though, Dr. Ullman decided to instead pursue a Ph.D. at MIT, where he began studying language under the mentorship of Dr. Steven Pinker, a renowned psychologist in the field of psycholinguistics. Although fascinated by his mentor’s cognitive approach to language, Dr. Ullman sought to study the neurobiological mechanisms of language. With guidance from both Dr. Pinker and another mentor, Dr. Suzanne Corkin, Ullman began to study patients with neurological disorders like Alzheimer’s, Parkinson’s, and Huntington’s diseases. These conditions primarily impaired different regions of the brain, offering a unique window into how language is processed neurologically. It was during this research that he developed the DP model.

Foundations of Memory: Declarative and Procedural Systems

To understand the DP model, it is necessary to understand the two different types of memory at play in the human (and animal) brain. Psychologists and cognitive neuroscientists have long classified learning and memory into two main categories: declarative memory and procedural memory. Declarative memory includes the learning and memory of information that is consciously known and can be voluntarily declared, such as hard facts and events. Procedural memory, which underlies the non-conscious learning and retrieval of skills, oversees learning activities like riding a bike or typing without looking down at the keyboard. Researchers have found that these two memory systems operate in distinct regions of the brain: while declarative memory primarily involves structures in the medial temporal lobe such as the hippocampus, procedural memory largely involves the basal ganglia. These two “classes” of memory are the focus of the DP model.

Formulating the Declarative/Procedural Model

The DP model posits that the brain encodes different components of language in different regions of the brain, corresponding to the pre-established memory systems associated with declarative and procedural memory. Previous research established two main components of language. First, there is the mental lexicon. This refers to the collection of words that exist in human mental repertoire — essentially, the connections the brain has made between a group of sounds (e.g., ca-ta-list) and their meaning (e.g., your favorite student magazine/a substance that increases reaction rates). The second component is mental grammar. This refers to the operations your brain performs — automatically in a first language — to communicate using the mental lexicon. For example, most native English speakers would not have to think twice before saying “I ate an apple” when referring to a past event, versus “I’ll eat an apple” in reference to the future.

As Dr. Ullman analyzed the results of his experimental study participants, he observed striking contrasts in the types of language impairments seen across different neurological disorders. While Alzheimer’s patients mostly displayed impairments in learning and retrieving words from their mental lexicon, Parkinson’s patients displayed more grammatical errors.

While pondering his work on a hike with his girlfriend, Dr. Ullman had an “aha moment” that eventually gave rise to what he now terms the DP model.

“Grammatical deficits may be due to problems with underlying procedural memory; lexical deficits with declarative memory,” he said. “Everything flowed from there. Of course, my current conception is very different — that’s science — but it started as a real epiphanic moment.”

Dr. Ullman realized that the contrasting results he observed in his study participants seemed to align with the characteristic patterns of neurodegeneration in both diseases: while Alzheimer’s involves degeneration of temporal-lobe structures, which are associated with declarative memory, Parkinson’s involves degeneration in the basal ganglia, associated with procedural memory.

Scientific and Practical Applications of The DP Model

Language, Disorders, and Evolution

The DP model has vast applications. Since first introducing his theory in 2001, Dr. Ullman has continued to study aspects of language neurocognition, such as second-language acquisition and bilingualism, language disorders, and other aspects of non-linguistic cognition, such as learning math and music.

Beyond illuminating the physical basis of language and its implications in the clinical setting, the DP model also highlights important principles of evolutionary biology that help scientists understand human linguistic development. “The procedural model is premised on the idea in evolutionary biology that new functions tend to rely on existing mechanisms that are repurposed for that function,” he said.

Ullman’s model both supports and is supported by this evolutionary perspective, suggesting that humans’ newly acquired capacity for language has organized itself in the context of the brain’s pre-existing learning structures — specifically, those underlying declarative and procedural memory.

Applications for Students: Optimizing Memory Science for Effective Learning

Dr. Ullman’s work has been influential in bridging the cognitive aspects of linguistic functions with their underlying neurobiology. His findings contribute not only to the study of language but also to broader research in learning and cognition. Many Catalyst readers — especially those studying biology, psychology, or other STEM courses — likely engage in content memorization on a daily basis. This objective, which heavily involves declarative memory, not only requires understanding general principles and details, but also associating them with words — for example, the term “mitochondria” with “powerhouse of the cell.”

The DP model provides insights that can aid intentional information learning. By characterizing the brain regions that facilitate this type of language memorization, such as the hippocampus and temporal lobe, researchers can leverage memory studies on these brain regions to identify optimal study strategies.

Currently, Dr. Ullman directs the Medical Neuroscience course for medical students at Georgetown Medical School. Drawing on research on declarative memory, he structures his own lectures to facilitate effective learning. Describing some of these strategies, Dr. Ullman notes the value of repetition of concepts, as well as the enhanced benefits of spacing out subsequent repetitions to promote information retention. Another key strategy he notes is “the testing effect,” also known as “retrieval practice,” which emphasizes the benefit of actively recalling information — as opposed to passively hearing — in long-term memory formation. Dr. Ullman makes sure to ask his students questions in real time to encourage active recall as well.

Conclusion: Integrating Language, Memory, and Mind

For students navigating lectures, lab reports, or exams, it can be easy to think of language and memory as separate and unrelated systems — Dr. Ullman’s research shows us just how connected they are. The DP model doesn’t just explain how the brain processes language; it also enables us to learn more effectively by understanding how we think, and even to appreciate how our brains have evolved to allow us to perform these functions. Whether memorizing metabolic pathways in biochemistry, synthesis reactions in organic chemistry, or an equation in physics, the brain is constantly utilizing declarative and procedural memory systems to make sense of the world. Dr. Ullman’s work helps to elucidate these functions and advance the work of neuroscientists, psychologists, educators, and even students by reframing our understanding of the biological machinery that underlies some of our most human capacities.

Sources

Brain and Language Lab – Research Overview. Georgetown University.https://brainlang.georgetown.edu/research (accessed May 14 2025).
Ullman, M. T. (2020). “The Declarative/Procedural Model.” In VanPatten, B. et al. (Eds.), Theories in Second Language Acquisition (Chap. 20). Georgetown University Press.https://georgetown.app.box.com/s/zjvbe8jdxpqcxmpf1xl6w2nu309ltt0i
Faculty Profile: Michael T. Ullman, Ph.D. Georgetown University.https://gufaculty360.georgetown.edu/s/contact/00336000014RrKZAA0/michael-ullman
Ullman, M. T. (2001). “The Neural Basis of Lexicon and Grammar in First and Second Language: The Declarative/Procedural Model.” Bilingualism: Language and Cognition.https://www.cambridge.org/core/journals/bilingualism-language-and-cognition/article/neural-basis-of-lexicon-and-grammar-in-first-and-second-language-the-declarativeprocedural-model/8AE8A67A3760AE58ED902E67A119B261
Ullman, M. T. (2001). “Contributions of Memory Circuits to Language: The Declarative/Procedural Model.” Cognition (Working Paper). (PDF) (accessed May 14 2025).