Odor Memories: The First Sniff Counts

A new study which combined associative memory tests with functional magnetic resonance imaging of the brain has identified a potential neural correlate of the special association that is formed when an odor is first paired with a visual object. Odor memories in humans are likely to be processed by a separate memory system with distinctive features that make odor memories different from visual or auditory memories [1,2]. Two peculiarities of odor memory fascinate scientists and poets alike: the vividness of odor-evoked memories; and the difficulty of remembering a smell. Odor-evoked memories are more vivid and emotional than memories evoked by stimuli in other modalities. Because the best known description of this phenomenon is found in Marcel Proust's novel In Search of Lost Time, this phenomenon is often called 'Proust effect'. On the other hand, memories of odors themselves are elusive. Some have argued that smells cannot be remembered at all — Vladimir Nabokov wrote in his debut novel Mary ''Memory can restore to life everything except smells''. Early research into the curious nature of smell memories relied on self-reported smell experiences and behavioral measures of odor memories. Functional brain imaging offers a new tool to objectively study how we remember odors by mapping activity in the brain while we remember. In this issue of Current Biology, Yeshurun et al. [3] report a study in which functional magnetic resonance imaging was used on subjects performing associative memory tests to study odor memories. Specifically they address the psychological phenomenon called 'resistance to interference' [4], or rather: what makes the first association of an odor with an object stronger than later associations of the same odor with a different object? Their results suggest that the hippocampus is involved in the formation of persistent odor associations. Resistance to interference of odor memories was first described by Lawless and Engen [5], who showed that over a two-week period the first of two associations to an odor were retained far better than the second. They concluded that ''If the storage system offers only one permanent link on a first-come first-served basis, subsequent learning will have little effect. The original associations will persist as long as they are unaffected by simple time-dependent decay processes''. In another set of experiments, Zucco [1] showed that the persistence of first-learned associations is specific for odor memories in an experiment that exposed subjects to olfactory, visual, or acoustic stimuli and then …