The overall goal of our research is to understand how memory for language is represented in the firing of neurons in the human medial temporal lobe. This research focuses on the differences in memory for abstract meaning of words vs. the details of their presentation, such as fonts and voices. Prior non-invasive studies in human subjects, as well as neurophysiological recordings in animals, suggest that the hippocampus is involved in the conscious recollection of linguistic stimuli. Existing theoretical explanations of memory posit that the hippocampus functions to quickly encode unique inputs and match stimuli to behavioral tasks, versus storing longer term memory for details of presentation. Both prior literature and preliminary recordings of single neuron activity in human epilepsy patients lead us, however, to a contrasting working hypothesis: that words presented and later repeated in the same voice (or typescript) should evoke strong MTL responses, relative to changed forms. To explore the implications of this hypothesis, we will record the activity of single neurons in the hippocampus of human epilepsy patients while the patients perform a continuous recognition memory (CRM) task. Our research is organized around three specific aims:
Specific Aim 1 : Determine whether single-neuron responses in the human MTL reflect the lags between items in continuous recognition memory for printed and spoken words.
Specific Aim 2. Determine how differences in episodic details, such as font or voice, may affect neural activity in the human MTL.
Specific Aim 3. Compare neural and behavioral responses when the CRM task engages memory either directly or indirectly. These experiments provide two major conceptual advances over previous single neuron recording in this population: 1) the CRM task is designed to dissect specific components of memory, addressing the differences between memory for abstract words and episodic details of their presentation;and 2) the use of spoken words will provide an older, and potentially more behaviorally relevant and evocative form of linguistic stimuli. Taken together, these experiments will allow a direct comparison of psychological theories of memory with neurobiology at the single cell level and will significantly advance our understanding of the neural representation of the uniquely human form of communication-spoken language.
An inability to correctly recall events is a hallmark of several major neuropsychiatric illnesses, such as Alzheimer's disease and senile dementia, and is a common comorbidity in neurological illnesses, such as epilepsy. Understanding how memory is represented by the fundamental computational units of the brain, single neurons, is a key step toward understanding and treating these illnesses.
| Valdez, André B; Papesh, Megan H; Treiman, David M et al. (2015) Distributed representation of visual objects by single neurons in the human brain. J Neurosci 35:5180-6 |
| Aur, Dorian (2012) A comparative analysis of integrating visual information in local neuronal ensembles. J Neurosci Methods 207:23-30 |
| Steinmetz, Peter N; Cabrales, Elaine; Wilson, Michael S et al. (2011) Neurons in the human hippocampus and amygdala respond to both low- and high-level image properties. J Neurophysiol 105:2874-84 |
| Fessler, A James; Treiman, David M (2009) Epilepsy and aggression: proceed with caution. Neurology 73:1720-1 |
