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

For anyone who speaks multiple languages, the experience of seamlessly switching between them can feel like operating two entirely different operating systems. This subjective experience, coupled with the occasional grammatical slip—such as a Spanish-English bilingual saying "I have 20 years" instead of "I am 20"—has long fueled a central question in neuroscience: does the brain build a separate grammatical engine for each language it learns?[1][3]

For decades, a prevailing hypothesis known as the "dual engine myth" suggested exactly that. The assumption was that learning a new language meant constructing a parallel, independent set of grammatical rules in the brain's gray matter, effectively compartmentalizing linguistic processing to prevent cross-contamination.[4]

A landmark study published today in the Journal of Neuroscience definitively dismantles that assumption. Researchers at New York University have provided high-resolution neural proof that bilinguals rely on a single, shared "grammatical engine" to process all the languages they speak, fundamentally redrawing our understanding of the multilingual brain.[1][2][3]

"Our research suggests that brains have a single grammatical engine that fuels all of the languages we speak—rather than separate engines for each one," explained Esti Blanco-Elorrieta, an assistant professor of psychology and neural science at NYU and the senior author of the study. The findings indicate that human language is built from neural computations that transcend any specific tongue.[3][4]

To capture the lightning-fast speed of human speech processing, the NYU team utilized magnetoencephalography (MEG). This advanced neuroimaging tool maps exact magnetic fields in the brain, allowing scientists to watch grammatical computations unfold down to the millisecond, providing a level of temporal resolution that older fMRI studies could not achieve.[4]

The researchers recruited Spanish-English bilinguals and subjected them to a "morphological stress test." During the MEG scans, participants were tasked with instantly transforming singular nouns into their correct grammatical plural forms across both languages—shifting the English word "boat" to "boats," or the Spanish word "barco" to "barcos."[2][4]

The empirical tracking data revealed an identical, language-transcendent neural template firing across both tongues. The exact same network of brain areas activated to adjust words to their grammatical context, regardless of which language the participant was actively speaking.[4]

To ensure the brain wasn't simply retrieving pre-memorized plural forms from a mental dictionary, the investigators introduced a crucial control mechanism: "pseudowords." These were completely fabricated words that followed the phonetic rules of the languages but held no actual meaning.[3][4]

Participants were asked to apply grammatical rules to these made-up words, such as pluralizing the nonsense word "paple." Remarkably, the brain's grammar-processing circuits engaged the exact same neural circuitry used for real words, proving that the brain was actively computing grammar rather than just looking up memorized vocabulary.[2][3]

This confirms that human grammar is executed as a highly reusable, universal computational loop. The brain operates on abstract principles that transfer across linguistic boundaries, deploying a universal language template rather than multiple language-specific rulebooks.[4]

The findings also offer a clear neurological explanation for why bilinguals sometimes mix up grammatical rules. These errors are not the result of two separate engines colliding or malfunctioning. Instead, they occur because a single, unified system is handling competing linguistic inputs simultaneously, occasionally applying the structural logic of one language to the vocabulary of another.[3][4]

This discovery builds on a growing body of evidence pointing toward a highly integrated bilingual brain. In February 2026, a separate study published in the Proceedings of the National Academy of Sciences (PNAS) found that bilinguals also share a single semantic system for understanding the meaning of words.[5]

That UC Berkeley study scanned Chinese-English bilinguals and found that 81 percent of the brain locations responding to meaning showed identical semantic tuning across both languages. A brain area tuned to family-related words in Chinese, for instance, was also tuned to family-related words in English.[5]

Taken together, the JNeurosci and PNAS studies provide a comprehensive, unified picture of the bilingual brain. Both the meaning of words (semantics) and the rules that govern their structure (syntax) rely on shared neural infrastructure, proving the brain is a highly efficient optimizer rather than a compartmentalized filing cabinet.[6]

The implications for language acquisition are profound. If there is one universal mechanism for language, it logically follows that learning subsequent languages becomes progressively easier because the foundational computational architecture is already firmly in place.[2][3]

Educators and cognitive scientists can leverage this insight to refine language instruction. Rather than teaching a second language as an entirely isolated and alien system, curricula could explicitly map new vocabulary onto the brain's existing grammatical templates, encouraging students to use their native language's engine as a bridge.[6]

However, researchers maintain transparent uncertainty regarding the absolute limits of this shared engine. The current NYU study focused on Spanish and English—two Indo-European languages that, despite their differences, share many underlying structural similarities.[6]

It remains an open question whether this identical neural overlap persists when a bilingual speaker navigates languages with radically divergent grammatical architectures, such as English and a polysynthetic language like Inuktitut, which constructs meaning through entirely different morphological rules.[6]

Furthermore, the neuroscience community continues to investigate whether this shared engine operates identically in individuals who learn a second language late in adulthood, compared to those who are raised in bilingual environments from birth.[6]

Despite these remaining frontiers, the latest neuroimaging data represents a paradigm shift. It validates the bilingual brain as a highly efficient, universal processor capable of weaving multiple linguistic threads through a single, elegant needle.[1][6]