Research shows that people have difficulty forgetting inferences they make after reading a passage, even when the information that the inferences are based on is later known to be untrue. This dissertation examined the effects of these inferences on memory for political information and tested if the credibility of the source of the correction influences whether people use the correction, or continue relying on the original information when making inferences. According to source credibility theory, there are two main factors that contribute to credibility, expertise and trustworthiness. Experiment 1 examined credibility as a function of both expertise and trustworthiness. The results from this experiment showed that having a correction from a source who is high on both factors significantly decreased the use of the original information. Experiment 2 examined credibility as a function of expertise. The Experiment 2 results showed no significant decrease in participants' use of the original information, if a correction came from a source that was simply more expert (but not more trustworthy) than another source. This finding suggests that source expertise alone is not sufficient to reduce reliance on the original information. Experiment 3, which examined credibility as a function of trustworthiness, demonstrated that having a highly trustworthy source does significantly decrease the use of the original information when making inferences. This study is the first to provide direct support for the hypothesis that making the source of a correction more believable decreases use of the original discredited information when making inferences.

No description available.

In this experiment it was investigated how scent affect the memory. Encoding information in the same context as retrieving it has been suggested to be beneficial for memory, earlier research has mostly explored how environmental contextual cues affects the memory. In this research the contextual cue was created by a presentation of a scent. The participants were presented with two lists of words, during the encoding of the first list all the participants were presented with a scent, half of the group was directed to forget the first list straight after encoding and the other half to keep remembering the list. For the second list no one was presented with a scent. In the retrieval of both lists half of each group was reinstated with the scent they were presented with at encoding and the other half without the scent (control group). The data were analysed with univariate ANOVAs and significant effects were followed up with independent-samples t-test. The results were that participants that were reinstated with the scent were thought to remember more than the others, however there was only a significant difference in the forget condition with reinstatement, where they remembered less than in the other conditions.

General intelligence is the exhibition of intelligent behavior across multiple problems in a variety of settings, however intelligence is defined and measured. Endemic in approaches to realize such intelligence in machines is catastrophic forgetting, in which sequential learning corrupts knowledge obtained earlier in the sequence or in which tasks antagonistically compete for system resources. Methods for obviating catastrophic forgetting have either sought to identify and preserve features of the system necessary to solve one problem when learning to solve another, or enforce modularity such that minimally overlapping sub-functions contain task-specific knowledge. While successful in some domains, both approaches scale poorly because they require larger architectures as the number of training instances grows, causing different parts of the system to specialize for separate subsets of the data. Presented here is a method called developmental compression that addresses catastrophic forgetting in the neural networks of embodied agents. It exploits the mild impacts of developmental mutations to lessen adverse changes to previously evolved capabilities and `compresses' specialized neural networks into a single generalized one. In the absence of domain knowledge, developmental compression produces systems that avoid overt specialization, alleviating the need to engineer a bespoke system for every task permutation, and does so in a way that suggests better scalability than existing approaches. This method is validated on a robot control problem and may be extended to other machine learning domains in the future.

Recent studies in rodents (McTighe et al, 2010; Burke et al, 2010) have suggested that forgetting is caused by the misidentification of novel stimuli as being familiar, matching the predictions of the representational-hierarchical model of Saksida & Bussey (2010). Here, we tested this idea in humans. Three groups of participants (young, healthy elders, elders at-risk for MCI) viewed novel and repeated stimuli in a continuous viewing task while their eye movements were recorded. According to the eye-movement based memory effect (Ryan et al, 2000) individuals make fewer fixations on items which are perceived as familiar. As interference increased, eye-movements directed to the novel stimuli declined, indicating these novel items were perceived as familiar. This effect was stronger in groups more vulnerable to interference (eg. at-risk elders). These results suggest that forgetting in humans, like rats, is driven by the misidentification of novel items as being familiar.

Recent studies in rodents (McTighe et al, 2010; Burke et al, 2010) have suggested that forgetting is caused by the misidentification of novel stimuli as being familiar, matching the predictions of the representational-hierarchical model of Saksida & Bussey (2010). Here, we tested this idea in humans. Three groups of participants (young, healthy elders, elders at-risk for MCI) viewed novel and repeated stimuli in a continuous viewing task while their eye movements were recorded. According to the eye-movement based memory effect (Ryan et al, 2000) individuals make fewer fixations on items which are perceived as familiar. As interference increased, eye-movements directed to the novel stimuli declined, indicating these novel items were perceived as familiar. This effect was stronger in groups more vulnerable to interference (eg. at-risk elders). These results suggest that forgetting in humans, like rats, is driven by the misidentification of novel items as being familiar.

No description available.

When people form episodic connections between memories that share a common retrieval cue, the tendency for those memories to interfere in later retrieval is often eliminated, and forgetting of the interfering information is reduced. For example, episodic integration protects memories from retrieval-induced forgetting (RIF), a phenomenon in which practicing retrieving some associates of a cue leads to the suppression of others that interfere with retrieval (Anderson, Green, and McCulloch, 2000). The purpose of this study was to determine whether semantic integration, as a result of pre-existing associations between practiced items and their unpracticed competitors, also moderates RIF. This research was motivated by the existence of many pre-existing associations between the practiced and unpracticed sets in one study that failed to replicate the RIF effect with item specific cues (Butler, William, Zacks, and Maki, 2001). It was hypothesized that pre-existing associations increase the implicit, semantic integration among the items, insulating them from inhibitory effects. The results were consistent with this hypothesis: when associations between the practiced and non-practiced sets were maximized, no forgetting was observed, however when such associations were minimized, there were reliable levels of RIF. The benefits of semantic integration were replicated across four experiments including one that used Butlers original materials and design. Furthermore, when Butlers items were simply re-arranged in order to minimize the associations and thus reduce semantic integration, the typical RIF effect was observed. Additional results revealed that the moderating effects of semantic integration are not mediated by explicit integration strategies.

"Occupying Memory: Rhetorical Studies for the 99%" revitalizes rhetorical memory by emphasizing memory's rhetorical production and non-declinable relationship to forgetting, the persuasive force of local genealogy, and the capacity of memory to spur invention and civic intervention. "Occupying Memory" performs its revival of memory through theorization of the contemporary Occupy Movement. The first chapter, "Becoming Activist," argues that memories are rhetorically produced, and supports this supposition by analyzing various activist practices, icons, and experiences. I consider the discursive production of memory through Occupy's practice of the "human microphone," and the imagistic production of memory through images such as the Guy Fawkes Mask. I also consider forgetting in the production of memory, and analyze how subjects are compelled to action through "forgotten" affects and traumas that drive one to compose self-narratives. "Giving an Account of One's Wealth," strives to develop a strategy for teaching writing called "im-personal writing," and employs Percentile Narratives from the Occupy Movement throughout its implementation. I analyze existing narratives from multiple theoretical perspectives, and focus on how students can consider the rhetorical production of their memories while avoiding the pitfalls associated with "personal writing" such as the quest for authenticity. "The Infinite Archive," considers how the binary opposition between so-called "live" and "technological" memory deconstructs, and avers that the digitization of memory is an instance of "hyper-extension" rather than "externalization." I consider multiple cases of such extension in the form of social media archives including Twitter, live streaming video, and viral memes. The problem of digital forgetting and networked multitudes is likewise engaged. "Stiller than Still" contends that (singular) bodies and specific living structures can function as monuments oriented toward the future. I argue that the type of memory such monuments produce is a "common" rather than "public" memory, one that entails resistance to state control, participatory democracy, and the preservation of difference. I also consider the nature of "common" forgetting in relation to affirmation. The text culminates with "Beginning(s)," as I consider how rhetorical memory and the Occupy Movement open onto the future, as well as the relation between memory, social movements, nostalgia, and hope. / text

For many years, El Salvador and Guatemala were submerged in brutal and bloody conflicts that cost the lives of tens of thousands. United Nations-brokered Peace Accords officially brought the years of violence to an end in 1992 and 1996, respectively. As the two countries slowly emerged from their respective Cold War-inspired internal conflicts, the question of what place the past would have in the present came to the fore. This dissertation explores the way past violence is talked about in the public sphere. It analyzes post-Peace Accords public discourse in both countries, with a particular focus on the issues of memory, forgetting, truth, reconciliation, and related terms. It examines the different tasks memory and truth were assigned in the Peace Accords, especially in relation to the truth/truth-like commissions created out of those accords, and in the years since, and looks at the language those who reject memory and truth use to oppose them. This dissertation argues that a common discursive framework exists in Guatemala that dictates that all sectors must insist on the importance of remembering the violence to prevent repetition. This is the human rights community's discourse, but it is one which even conservatives who wish for forgetting must repeat. Conservatives can only promote forgetting within the limits of this discursive framework, and they do so by talking about amnesty, perdón (pardon/forgiveness), and reconciliation. The situation in El Salvador is different. There is no common discursive framework that demands memory to prevent repetition and promote reconciliation. Rather than this, conservatives openly insist on amnesty and amnesia, while the human rights community insists on truth and memory. The discursive battle between forgetting and truth is El Salvador's discursive framework. Yet talking about memory, truth, reconciliation, and related topics leaves space to promote different truths, memories, or narratives of the past. This, indeed, is precisely what happens in both countries as different sectors actively promote their own truth, memory, or narrative, especially at moments of rupture or when their truth or discourse is challenged, as in 2012 when Salvadoran president Mauricio Funes asked for perdón for the El Mozote massacre and during Guatemala's 2013 genocide trial. Running throughout the discussion about discourse and discursive frameworks is a critique of the insistence on the existence of one truth, memory, or narrative of the past. This is the foundation on which truth and truth-like commissions are built. Yet rather than focusing on the truth of the past, this dissertation argues that the process of openly talking about the past and sharing truths and experiences will do more to contribute to reconciliation and non-repetition than insisting that there is and can only be one truth and that everyone must embrace it.