Can We Create New Senses for Humans? Inside the Science of Sensory Expansion

Yes. Researchers have already given people new senses in the lab. Blind participants have read printed letters through a tongue stimulator. Deaf volunteers have perceived sound through fingertip vibration. Healthy subjects have learned to feel an invisible magnetic north. The senses were not biological, but the brain treated the incoming data as sensory information. The deeper question is not whether we can build new channels, but how far the nervous system will go to interpret them.

The mechanism: how the brain adopts foreign data

Every sense begins with transduction, the conversion of one kind of physical energy into electrochemical signals. Photoreceptors turn light into voltage. Hair cells in the cochlea turn pressure waves into spikes. Touch receptors in skin turn deformation into spikes. The brain never receives light or sound directly. It receives spike patterns and learns, over development, which patterns mean what.

This is why sensory substitution works. A camera can capture an image, a computer can convert brightness into a grid of vibration or electrical pulses, and a stimulator can deliver that grid to the tongue or skin. The peripheral hardware changes entirely. The cortex still gets spikes. With training, those spikes acquire meaning.

The key structure is the sensory cortex, which is organized by modality but not rigidly locked to it. In people blind from early life, the visual cortex responds to touch and sound. In deaf people, auditory cortex regions respond to sign and vibration. This cross-modal plasticity means unused territory can be recruited. A new channel does not need to be visual or auditory in origin. It needs to be informative, consistent, and behaviorally relevant.

Two features determine whether a signal becomes a usable sense. First, it must be congruent: sensor motion should match the information it carries. A camera pointed left should produce leftward activity in the map. Second, it must be contingent: the user's own actions must change the input. Contingency is what lets the brain learn a sensor as part of the body rather than as an external tool. A 2020 review in Nature Neuroscience described this as the difference between watching a screen and wearing a sense.

The plain version: the brain does not care where signals come from, only whether they are reliable and tied to what you do.

Quantified context: what has actually been demonstrated

The most studied device is the BrainPort, a tongue display with 100 to 400 electrodes. In trials, blind adults have used it to recognize printed letters, locate doorways, and catch rolling balls. Training is short by sensory standards. Some participants reached basic letter recognition within hours, and performance improved over days to weeks.

Vibrotactile hearing aids, which convert sound into skin vibration, have let deaf users detect environmental sounds and some speech rhythms. Speech comprehension remains poor except with training, and outcomes vary widely. Cochlear implants are the strongest counterexample: they are not a new sense but a restored one, with over 1 million recipients worldwide and speech understanding in quiet for many adults. They show that direct electrical stimulation of a sensory nerve can produce usable perception when the peripheral organ is gone.

Magnetic sensing has been tested in healthy volunteers using a belt that vibrates according to compass heading. After several days, users reported a persistent felt sense of direction, not just a buzzing cue. A frequently cited 2015 study in eNeuro found that blind participants using the BrainPort showed visual cortex activation during tongue stimulation, supporting the recruitment idea.

Timeline of sensory expansion research

Those dates and numbers should be read as research milestones, not product timelines. No non-restorative new sense has received regulatory approval as a medical device. The BrainPort is investigational for vision substitution. Most augmentation devices are consumer or lab tools, with prices ranging from under USD 100 for open-source haptic kits to thousands for research systems. Treat any specific price as volatile.

Edge cases: where the model bends

Children adapt faster than adults. The developing cortex is more plastic, and children with sensory substitution training show faster and more complete integration. This is consistent with critical-period biology, but it also raises ethical questions about implanting experimental channels in minors.

People with congenital blindness versus acquired blindness differ. Congenital blindness is associated with stronger cross-modal recruitment, but acquired blindness can also produce reorganization. The direction of that difference is not fully settled, and studies disagree on which group performs better on specific substitution tasks.

Sensory conflict is a real failure mode. If a new channel disagrees with an existing sense, the brain may suppress or reject the new signal. Congruence and contingency reduce this, but they do not eliminate it. Users report transient dizziness or confusion during early training, especially with vestibular-like signals.

There is also a ceiling. Adding a channel does not automatically add bandwidth. Attention is limited, and a new sense competes with existing ones for cortical resources. Some users show interference: their performance on a primary task drops while they use the new channel.

Counterintuitive detail: a new sense can feel like a new sense without being one biologically. Users of the magnetic belt described a direct perception of direction, yet the underlying signal was still vibration on the skin. The brain relabeled the source. This is a professional caveat for anyone claiming a device "creates" a sense rather than "delivers" information the brain learns to interpret.

Misconceptions worth correcting

The biggest myth is that a new sense must be wired into a dedicated cortical area. It does not. The brain assigns meaning to patterns based on reliability and relevance, not on anatomical destination. A second myth is that more electrodes mean better perception. Resolution matters, but so do training, congruence, and task design. A 400-electrode tongue array is not automatically better than a well-designed 100-electrode system.

Another misconception is that sensory augmentation is plug-and-play. It is not. Learning a new channel resembles learning a language or an instrument: consistent practice over days to weeks, with feedback. Without feedback, the signal stays noise.

A related error is confusing restoration with augmentation. Cochlear implants restore hearing to people who lack it. A magnetic belt adds a sense to people who do not need it. The regulatory, ethical, and scientific standards differ.

Red flags to watch for in product claims: A device claims to "expand human perception" without peer-reviewed training data. No description of congruence or contingency design. No reported effect sizes, only testimonials. Claims of immediate use with no learning period. Medical claims without FDA clearance or equivalent regulatory status. Prices and availability presented as fixed when they are not.

Practical actions: how to evaluate or try sensory expansion

If you are curious, start with a low-stakes, reversible, open protocol. The goal is to test whether your brain will adopt the signal, not to buy a product.

Choose a channel with clear congruence, such as a camera-to-vibration map or a compass-to-belt map. Commit to daily practice of 15 to 30 minutes for at least two weeks before judging. Use active tasks, such as finding a target or navigating a route, not passive watching. Include immediate feedback so errors are correctable. Track performance in a simple log: accuracy, time, and subjective effort. Stop if you experience persistent dizziness, headache, or attention interference, and consult a clinician if symptoms continue.

Thresholds matter. Research training periods typically run from a few hours to several weeks. If a protocol claims results in under an hour with no feedback, be skeptical. If a device costs more than USD 1,000 and has no published effect sizes, ask for the data before buying. Reconcile thresholds across sources by asking what outcome is measured: detection, discrimination, or navigation. A device can improve detection while leaving discrimination unchanged, and both are legitimate results.

For clinicians and researchers, the practical rule is to separate feasibility from efficacy. Feasibility is whether a signal can be delivered and learned. Efficacy is whether it improves function in real tasks. Most new-sense studies demonstrate feasibility. Fewer demonstrate durable efficacy outside the lab.

Limitations: what we still cannot do

We cannot yet create a sense that is as rich as vision or hearing. Bandwidth, resolution, and dynamic range are far lower. We cannot guarantee adoption in adults, and we cannot predict who will adapt quickly. We do not have long-term safety data for chronic stimulation of new channels, particularly electrical stimulation of skin or tongue over years.

Regulatory status is another limit. No new-sense device is approved for routine clinical use beyond restoration. Insurance does not cover augmentation. Costs and availability change, so verify current status directly with the manufacturer or a clinician.

The deepest limitation is conceptual. A new sense is not a new window onto reality. It is a new data stream the brain must learn to trust. That learning is real, measurable, and reproducible, but it is also slow, individual, and constrained by attention and plasticity. The honest answer to whether we can create new senses is yes, within limits: we can deliver new information channels and train the brain to use them, but we cannot yet manufacture perception at the fidelity of biology.