GenjiKō - Sequential Scent Research Background
Imagine a world where scent isn't just something you smell, but something you can "see" and remember. This is the premise of Genji-kō (源氏香), a sophisticated social game from Japan's Edo period where participants "listen" to and identify different incense scents. This game, and the elegant visual system it uses, became the foundation for my research into how we perceive and recall olfactory information.
From Scent to Symbol: The Genji-mon System
At the heart of Genji-kō is the Genji-mon (源氏紋) system, a unique symbolic language that represents scent sequences. Players are presented with five incenses, and as they identify the different smells, they record the unique sequence using one of 52 distinct patterns. These patterns, derived from a set of five primitives, are more than just a scoring system; they are a form of synesthesia, linking the abstract experience of scent to a tangible visual form.
These patterns aren't random; they serve as "scent bookmarks" for The Tale of Genji (源氏物語), the classic Japanese novel by Murasaki Shikibu. Just as the book's characters use scent to express themselves, the Genji-mon patterns create an inter-subjective language of fragrance.
The Science Behind Scent and Memory
My research project, Genji-Go, investigates how this multimodal approach—linking scent to a visual system—impacts our cognitive processes. My project's foundation is built on two key scientific principles:
Cognitive Load Theory (CLT): This theory suggests that our ability to recall olfactory information declines as the complexity of the scent sequence increases. The more smells you have to remember, the harder it gets.
Weber-Fechner Law: This law states that our perception of a stimulus (like scent) is relative, not absolute. We notice the difference between two scents, but not their individual absolute intensity.
Building on these ideas, my research aims to answer a fundamental question: How does a tangible, multimodal interface that maps sequential olfactory information to visual patterns affect a user's memory encoding, recall accuracy, and cognitive load?
My hypothesis is twofold: first, that a visual representation will reduce the cognitive load of a purely olfactory task, making it easier to remember the scent sequence. Second, that this multimodal interface will create a stronger, more lasting memory trigger.
Genji-Go: A Modern Scent Visualizer
To test my hypothesis, I'm developing Genji-Go, a physical-digital system that serves as a multimodal scent visualizer to improve memory recall. As users experience a sequence of scents, the system generates corresponding visual patterns from the Genji-mon system. This allows users to "see" the scent sequence, providing a tangible, visual anchor to help them encode and recall the olfactory information, thereby reducing cognitive load.
Initial playtests of Genji-Go revealed promising results. Participants quickly learned the Genji-mon patterns and found them accessible, even as newcomers to the game. One player, EK, noted, "I am able to learn the patterns quickly and visualize new scent sequences using those patterns." This feedback suggests that the Genji-mon system, with its elegant emergent complexity, could indeed be a powerful tool for enhancing memory and perception.
This research, and the creation of Genji-Go, is dedicated to the legacy of Masataka Hata, the 12th Generation President of Shoyeido, a master of incense whose life's work deeply inspired this project. As Joi Ito once wrote, his influence "will continue burning gently in our hearts, like the finest incense he so lovingly shared with the world."
Genji-Go raises a compelling final question: Can AI learn how to play Genji-kō and truly understand human scent perception? The future of scent, memory, and technology smells intriguing.
