Vol 443 | Issue no. 7107 | 7 September 2006 AUTHORS

inclusions, the pockets of melt trapped by crys- Abstractions MAKING THE PAPER tals deep below Earth’s surface. By examining FIRST AUTHOR Kathy Cashman various inclusions trapped at different times Studies of the visual during the ’s ascent, they attempted to system have revealed a piece together how magma changes as it works lot about how we process monitoring heats up with its way to the surface. simple features of an new magma knowledge The work relied on 25 years’ worth of sam- object such as its shape, ples from Mount St Helens in Washington and colour and movement. Volcanologists still aren’t particularly adept at additional samples collected over a period of Much less is known about working out exactly when volcanic eruptions five years from the Kamchatka Peninsula, a vol- how the brain encodes the meaning, or will occur, how they happen or how long they canic hotbed in eastern Russia. Cashman and category, of visual stimuli. On page 85 of this are likely to last. Now, Kathy Cashman at the Blundy analysed selected inclusions with an issue, David Freedman, from Harvard Medical in Eugene and her col- instrument known as an ion microprobe. This School, reports the identification of monkey leagues have found a geological clue about the measures the water content and the concentra- brain cells that they believe allot meaning to heating of magma (see page 76). This could tion of different elements by blasting away part objects. These cells are located in the lateral help modellers better describe the several- of the sample and running it through a mass intraparietal (LIP) area, a brain region that uses visual information to help monkeys kilometre journey this molten rock makes to spectrometer. By relating the inclusions to the make decisions such as to move their eyes to Earth’s surface. temperatures at which they were preserved, or from an object. Freedman teamed up with How magma rises to the surface depends on these researchers discovered not only that a John Assad, also at Harvard, to show that LIP a number of factors. Many are water- significant amount of heating occurs during neurons can code for a specific category and rich, and as they rise the external pressure this process, but that magma crystallization is a retain that information during a delay period. decreases and water is lost in the form of gas. relatively fast phenomenon, taking years rather The information is stored in the monkeys’ This water loss changes the properties of the than centuries. short-term memory. Freedman spoke to melt such that crystals form and the magma’s Although her work is of help to modellers, Nature about how the brain assigns meaning viscosity increases. Cashman much prefers to be out in the field to moving visual targets. Such factors are of great interest to Cashman collecting samples rather than crunching num- and her long-time collaborator Jon Blundy, a bers on a computer. One of her favourite tasks How did you come to study categorization? petrologist at the , UK. is retrieving samples from active volcanoes that I’m driven by the desire to answer the Last year, they published work tracking the are percolating (as opposed to exploding) with question of how the brain makes sense of our decompression, gas loss and crystallization of lava. Pieces of pasty lava roll down the side of visual surroundings. We have all these objects magmas. While evaluating their results, they the volcanic dome — the mound that is formed around us — such as tables and chairs — and realized that something else was happening. when lava is so viscous that it cannot flow out we know what to do with them, but how? We When a crystal forms, it gives off heat. “We of the volcanic crater. Cashman and her col- taught monkeys to watch tiny dots drifting in were actually just fussing with the samples leagues run over to blocks that have rolled a any one of 12 directions and then categorize and this sort of fell out,” Cashman says. “We safe distance from the dome, knock off a piece those directions into two groups of six. all of a sudden said, ‘Oh my God, look at what with a hammer, then plunge it into the nearby Why use direction of motion rather than, we’re seeing!’” snow to cool. say, chairs? Like molasses, magma gets runnier as it heats Building on field work and geological analy- Partly because we already know a lot about up, Cashman explains. Tracking magma move- sis, Cashman and Blundy continue to dissect how the brain processes visual-motion ment and temperature should help researchers the finer points of magma and eruptions. They patterns — the anatomy and physiology are better understand volcano-monitoring data. now plan to investigate how the ascent, degas- well understood. And also because so much “Although people had theorized, ‘Oh yeah, sing and crystallization of these small ‘pack- of our behaviour involves reacting to events latent heating should be happening,’ we were ets’ of magma vary with eruption style — why in our visual world. If something is moving actually able to document it,” says Cashman. magma travels quickly or slowly and what towards you, do you grab it or move out of its “I try to pose my research questions such that makes it explode through the surface or form way? We need to understand motion to cross the findings can be used to improve both the a dome. They want to know, says Cashman, not a busy street or recognize whether someone models and ideas about how to monitor volca- only what makes magma come to the surface is waving hello or motioning for us to go away. noes,” she adds. She and Blundy study volcanic but what keeps it coming. ■ Even simple animals such as flies can detect motion and respond appropriately.

Will this work shed light on memory loss? KEY COLLABORATORS People often tell me they hope I will make a In 1994, two Boston College “Now, it’s a pure collaboration culture almost discourages it,” pill to improve memory. That is probably a chemists with neighbouring in which two synthetic chemists he says. Early on, colleagues long way off, but we are at least zooming in offices discussed their go after complex problems warned Snapper that working on areas that have a role in storing long-term respective scientific together,” says Hoveyda. with a more senior researcher memories. This will hopefully lead to a better frustrations. Amir Hoveyda Most recently, they found a could hurt his career. understanding of amnesia and Alzheimer’s, in which people often show impaired was having trouble identifying more environmentally friendly But their partnership has learning, memory and recognition. cleaner, more efficient catalysts way to synthesize small been fruitful; the two have in a timely manner, and Marc molecules for therapeutic use received joint grants from the You’ve studied aspects of vision since you Snapper was seeking uses by skipping several waste- National Institutes of Health were in college — why the fascination? for the peptides he had been producing steps (see page 67). since 1997 and have published As a student, I attended a course about studying. They soon agreed to Hoveyda says the synthetic about 20 papers together. how the brain makes sense of sensations. draw on Snapper’s synthesis organic chemistry field has “The reason this collaboration I realized these were questions I had been background and Hoveyda’s traditionally been wary of two has been so successful is that thinking about for years, such as: “Does your catalysis experience. The principal investigators sharing neither of us cares who gets the red look like my red?”. ■ relationship quickly evolved. credit on single papers. “The credit,” Hoveyda adds. ■

xvii © 2006 Nature Publishing Group