Dr. Denise Herzing has spent 40 years studying dolphin communication in The Bahamas. She founded the Wild Dolphin Project in 1985, recording underwater video and sound of Atlantic spotted dolphins in their natural habitat. Her goal has always been understanding how these animals communicate with each other and whether they might have something approaching language.
About 15 years ago, her team introduced something unusual to their research: CHAT, which stands for Cetacean Hearing and Telemetry. This underwater computer system creates synthetic whistles that don't exist in the dolphin's natural sound repertoire. Instead of trying to decode and mimic dolphin communication, they created entirely new signals. Signals neither species used before.
The reasoning behind this approach might seem counterintuitive. If you want to communicate with dolphins, why not learn their language? Dr. Herzing's answer is simple: because you might accidentally say something offensive! You might communicate the wrong thing entirely. You don't know what you don't know about their system. So instead, you create neutral ground. You build something new together.
Dolphins aren't humans. Associations aren't trying to communicate with marine mammals. But there might be something here worth considering about the value of starting fresh rather than forcing one party to adapt to systems that were never designed with them in mind.
The CHAT System Approach
CHAT is essentially a wearable underwater computer. The team uses it when dolphins are already playing with them, never interrupting natural dolphin behavior. The system has four simple signals, synthetic whistles created specifically for this interaction, that correspond to objects dolphins like to play with: sargassum seaweed, rope, and scarf.
The whistles are synthesized, meaning they're created outside the dolphin's natural repertoire. This was intentional. Dr. Herzing didn't want to risk using sounds that already meant something in dolphin communication. Imagine trying to communicate with someone in a language you barely understand and accidentally saying something rude. You want to avoid that scenario entirely.
The goal was creating mutual signals both species could use together. If a dolphin mimics the synthetic whistle for sargassum, they get the object and everyone plays together. The team gets to interact with the dolphins. The dolphins get to play with interesting objects and engage with the humans. The motivation comes from social interaction itself, not food rewards or physical touch.
CHAT uses off-the-shelf equipment now, including a Pixel phone, all contained in a small waterproof box researchers wear while swimming. The technology has evolved significantly from the early versions, but the core concept remains: create shared vocabulary rather than decode an existing language system.
Why Not Just Learn Their Language?
Creating mutual signals seems easier than decoding an entire existing language. Dr. Herzing offered a human analogy. Imagine two people from different cultures who don't speak each other's languages. One calls a phone "phone" and the other calls it something completely different. Instead of one person learning the other's full language, they could agree to call it "tutu." Now both understand this new word, even though they each have their own complete languages.
That's a simple interaction. You couldn't have complex conversations with just six shared words. But it demonstrates something valuable: you're reaching out, trying to communicate, and empowering the other party to communicate back. For simpler interactions, that might be sufficient.
This pattern exists in nature. Different dolphin species that interact with each other have mutual signals they use when together. They don't necessarily understand each other's complete communication systems, but they have critical signals for coordination. Birds and monkeys show similar patterns when interacting with other species.
The historical approach to animal communication has involved keyboard interfaces. Dolphins have had acoustic interfaces similar to what primates might have with physical keyboards. These experiments aimed to teach animals human-created symbols and see if they could use them meaningfully. CHAT follows this tradition but adds the element of creating something entirely new rather than adapting human language.
From a research perspective, creating mutual signals offers another pathway into understanding dolphin minds. Dr. Herzing's primary work focuses on decoding natural dolphin communication (that's where her 40-year data set comes in) and where AI tools like DolphinGemma are now helping to reveal patterns. But CHAT represents a secondary approach. A different way to explore the question of dolphin intelligence and communication capacity.
The Unexpected Discovery
CHAT hasn't produced the language breakthrough Dr. Herzing hoped for. Fifteen years into the project, they're still working on it. Georgia Tech researcher Thad Starner and his students thought it might be a quicker project. It hasn't been.
But something interesting happened anyway. While studying how dolphins mimicked the synthetic whistles, they discovered the dolphins were mimicking in ways the team hadn't seen before in their natural sound repertoire.
That discovery now requires going back through decades of basic sound data to search for those patterns in natural dolphin communication. Maybe those patterns were always there and researchers just weren't looking for them. Maybe the CHAT interaction prompted dolphins to use communication methods they don't typically employ. Either way, the experiment revealed something unexpected.
Science often works this way. You design an experiment to test one hypothesis, and the results point you toward something completely different. CHAT might not have created the mutual language system researchers envisioned, but it exposed communication patterns they would have otherwise missed. The "failure" generated valuable information.
Dr. Herzing also notes they've been incorporating machine learning and AI into CHAT work. They can now prompt DolphinGemma with the synthetic whistles and see different ways dolphins might mimic them. The system generates possibilities, maybe a hundred different variations. Then researchers review and decide which ones they would accept as mimics. Previously, students had to manually generate these variations, stretching and compressing sounds, adjusting pitch. AI speeds up that process considerably.
What This Might Mean for Associations
Dolphins are not members. The parallel isn't perfect. Members are humans with complex motivations, decades of professional experience, and clear expectations about what associations should provide. Dolphins are wild marine mammals playing with seaweed and rope.
But consider how often associations force members into existing systems because "that's how we've always done it."
Member portals designed around staff workflow rather than how members actually want to access information. You organize content by your internal department structure. Members have to learn your organizational logic to find what they need.
Feedback mechanisms that ask questions associations want answered rather than questions members want to ask. Survey design reflects your strategic planning cycle, not the issues keeping members up at night.
Programs structured around organizational capacity instead of how members actually learn or work. You offer webinars on Tuesday afternoons because that's when your education team has availability, not because that's when members can attend.
Conference schedules that prioritize sponsor obligations and keynote speakers over the sessions members actually need. The educational content gets squeezed into whatever slots remain after the networking lunch and exhibit hall hours are set.
Credentialing requirements built around what's easy to track and administer rather than what actually advances professional competency.
These aren't necessarily wrong decisions. They often make practical sense given organizational constraints. But they do represent asking members to adapt to systems that perhaps weren't designed with them as the primary consideration. They're your language, not mutual signals created together.
What if instead of making members learn your systems, you co-created new approaches? What if you started from "what do members actually need" rather than "how do we fit member needs into our existing structure"?
Meeting on Neutral Ground
This doesn't mean abandoning all existing systems and starting from scratch. Dr. Herzing hasn't abandoned her 40 years of natural dolphin communication data to focus exclusively on CHAT. Both approaches coexist. The traditional work continues alongside the experimental work.
But it does mean recognizing when you're asking members to do things your way because it's easier for you, not better for them. It means catching yourself when the justification for a process is "we've always done it this way" rather than "this serves members effectively."
Starting fresh sometimes reveals solutions neither party would have designed alone. Dr. Herzing could have kept trying to mimic dolphin sounds with increasing sophistication. Instead, creating something new together opened different possibilities. The dolphins responded in ways that exposed communication patterns researchers hadn't seen before.
The willingness to say "maybe our way isn't the best way" creates space for innovation. It signals to the other party that you're genuinely trying to meet them rather than requiring them to meet you. That changes the dynamic.
For associations, this might look like:
Asking members to co-design processes rather than requesting feedback on processes you've already built
Creating pilot programs that intentionally break from standard procedures to test whether new approaches work better
Admitting when legacy systems serve organizational needs more than member needs, then actually doing something about it
Letting members choose how they want to engage rather than funneling everyone through identical paths
Building flexibility into requirements instead of rigid structures everyone must follow regardless of circumstance
The CHAT system gives dolphins choice. They can participate or not. They can leave whenever they want. The interaction happens on terms that work for both species. Associations could ask themselves whether members feel they have similar choice, or whether participation often feels like compliance with predetermined systems.
What Gets Created Together
Dr. Herzing's CHAT system remains a work in progress after 15 years. It hasn't cracked the dolphin language code. The mutual signals haven't created the communication breakthrough she initially hoped for.
But it revealed communication patterns she wouldn't have found otherwise. It demonstrated that meeting another species on neutral ground creates different possibilities than forcing them to adapt to human systems. It showed that motivation can come from the interaction itself rather than requiring external rewards.
Your members aren't dolphins. But they might benefit from associations willing to question whether "the way we've always done it" is actually the best way, or just the most familiar. They might respond differently if you stopped asking them to learn your language and started building mutual signals together.
Sometimes the experiment doesn't work as planned. Sometimes you spend 15 years on something that hasn't produced the results you wanted. But if you're paying attention, you discover things you weren't looking for. Things that only become visible when you're willing to start fresh rather than defend existing systems.
The dolphins keep coming back to play with Dr. Herzing's team. That suggests something about the interaction is working, even if it's not working the way researchers originally envisioned. Maybe that's the real lesson: success doesn't always look like you expected, but you only find out what it looks like if you're willing to try something new.
.svg)
