Bilingual Brains Rely on a Single 'Grammatical Engine' for All Languages, Study Finds

For decades, cognitive scientists have debated a fundamental question about the bilingual mind: when a person speaks two languages, does their brain build two separate grammatical rulebooks, or does it rely on a single, shared system? A new study published in the Journal of Neuroscience has definitively answered this question, fundamentally redrawing our understanding of human language.[1][4]

The research, led by neuroscientists at New York University, proves that multilingual speakers do not possess separate grammatical engines in their gray matter. Instead, the brain utilizes a single, universal computational loop to process the grammar of every language a person speaks, seamlessly applying the same neural architecture across different vocabularies.[2][3]

Esti Blanco-Elorrieta, an assistant professor of psychology and neural science at NYU and the study's senior author, noted that the findings challenge long-held assumptions. The research suggests that human language is built from neural computations that transcend any one specific tongue, rather than relying on isolated, language-specific modules.[1][3]

To prove this, the research team utilized magnetoencephalography (MEG), an advanced neuroimaging technique capable of tracking brain activity on a millisecond-by-millisecond basis. They monitored Spanish-English bilingual speakers as they performed real-time grammatical transformations, capturing the exact moment the brain applied a grammatical rule.[3][4]

Participants were asked to listen to singular words in both English and Spanish—such as "boat" and "barco"—and instantly produce the plural versions, "boats" and "barcos." The MEG scans revealed that the exact same neural patterns fired to support grammatical processing in both languages, indicating a deeply shared computational architecture.[3][4]

A critical challenge in linguistics is proving that the brain is actively calculating grammar rules on the fly, rather than simply pulling memorized, pre-pluralized words from a mental storage locker. To bypass this, the researchers introduced "pseudowords"—completely fabricated terms like "paple" that the participants had never encountered before.[2][5]

When bilingual participants were asked to pluralize these fake words, the MEG scans showed the exact same shared neural engine engaging. Because the participants could not rely on memory for a word that did not exist, the experiment proved that the brain possesses an abstract, reusable grammatical formula that it can instantly stamp onto any new vocabulary.[3][4]

This shared architecture perfectly explains a common phenomenon among multilingual individuals: grammatical overlap. It is not uncommon for a bilingual speaker to mistakenly apply the rules of one language while speaking another—for example, a native Spanish speaker saying "I have 20 years" instead of "I am 20" when speaking English.[1][3]

Previously, some theorized this was a sign of the brain accidentally crossing the wires between two separate language systems. The NYU study proves the opposite is true. These slip-ups happen precisely because the brain runs all vocabulary through the exact same computational loop, occasionally causing a rule from one language to temporarily map onto another.[2][6]

The findings offer profound implications for how humans learn and communicate. By demonstrating that grammar is a reusable computation—a universal language template—the research suggests that learning a second language does not require building a new cognitive infrastructure from scratch.[3][5]

Instead, language learners are simply feeding new vocabulary into an existing, highly efficient engine. This aligns with broader neurobiological theories proposing that language selection operates through activation-based mechanisms without requiring the active suppression of a non-target language.[4][5]

While the current study focused on Spanish and English—two languages that share some structural similarities—the researchers believe this shared neural mechanism likely transcends any specific language pairing. The evidence points toward a universal neurobiology of human communication, where a single engine powers the vast diversity of human speech.[1][6]