1985 |
Miller, Ralph R. |
R01Activity Code Description: To support a discrete, specified, circumscribed project to be performed by the named investigator(s) in an area representing his or her specific interest and competencies. |
Extra-Acquisition Assoc Determinants of Performance @ State University New York Binghamton
Associative performance deficits are frequently attributed to acquisition failure with little empirical justification. Recent demonstrations that restorative treatments (often reminders designed to exclude further acquisition concerning the target information) can reverse many associative performance deficits suggest that these deficits are due to retrieval failure or response competition. Using rats in classical conditioning paradigms due to the high degree of stimulus control afforded, we will continue this research with the ultimate goal of contributing to an information processing model relevant to normative human behavior. Half the proposed research will focus on effects of context. In addition to direct context-US associations and contextual potentiation of retrieval of associations between the nominal CS and US, it appears that context servies as an associative baseline against which an organism evaluates a nominal CS. Prior research on this last issue has confounded training context with test context. We plan to independently vary the associative strength of these two contexts. Preliminary work indicates that the associative strength of the test context is used for comparison with the target CS. Noting that 0 less than or equal to P(US/CS)less than or equal to 1, which suggests that all associations between CSs (including contexts) and USs are neutral or excitatory, the approach leads to the conclusion that conditioned inhibition is not a property of an association, but only the consequence of testing a CS in a context more excitatory than the CS itself.--The other half of the research will use reminder techniques to determine if latent inhibition, one-trial overshadowing, and one-trial blocking can be added to our list of associative performance deficits that do not arise entirely from impaired acquisition. Additional experiments will determine if our recent observation of recovery from multitrial overshadowing and blocking depends equally upon all the reinforced compound trials or only upon the first few; i.e., are overshadowing and blocking on early compound trials an acquisition failure but on later compound trials a retrieval failure? Another series of studies will determine, when A blocks or overshadows X, if recovery of responding to X is the consequence of a direct X-US association or of X-A and A-US associations. We also plan to test Bolles' hypothesis that animals differentiate contiguity from contingency, with both being learned but only the latter controlling behavior.
|
1 |
1988 — 1999 |
Miller, Ralph R. |
R01Activity Code Description: To support a discrete, specified, circumscribed project to be performed by the named investigator(s) in an area representing his or her specific interest and competencies. |
Extraacquisition Associative Determinants of Performance @ State University New York Binghamton
Acquired performance is not merely a direct reflection of te associative strength of elucidating stimuli. Research using Pavlovian conditioning has demonstrated that other cues present during testing can modify the behavioral impact of eliciting stimuli. We propose to investigate properties of behavioral modifiers called 'occasion setters.' which appear to work by informing the subject of whether or not a subsequent conditioned stimulus is going to be reinforced on that trial. Limited recent evience has suggested that rules for processing of information about occasion-setting are similar to those for simple Pavlovian conditioning. As considerable knowledge of processes governing Pavlovian conditioning already exists, we propose to extend this analogy by seeking additional parallels in occasion setting to well establish Pavlovian processes, thereby further illuminating the nature of occasion setting. Although occasion setting and simple conditioning possible obey parallel rules, evidence suggests that a stimulus's occasion- setting attributes and its Pavlovian excitatory attributes are independent. That is, changing either Pavlovian or occasion-setting attributes of a stimulus does not appear to alter the stimulus' occasion-setting or Pavlovian attributes, respectively. We have demonstrated modulation of the meaning of occasion setters by what we call secons- and third-order occasion setters, and we now propose to investigate the independence of simple excitatory conditioning, first- order occasion setting, and second-order occasion setting. We anticipate both transfer and blocking will occur within a common level of behavioral control, but not between levels of control. Additional studies will examine the role of occasion setting in instrumental discrimination tasks and sequence learning, and the value of hierarchical memory models in explaining occasion setting.
|
1 |
2000 — 2009 |
Miller, Ralph R. |
R01Activity Code Description: To support a discrete, specified, circumscribed project to be performed by the named investigator(s) in an area representing his or her specific interest and competencies. |
Associative Determinants of Performance @ State University New York Binghamton
DESCRIPTION (provided by applicant): For the last 22 years, we have demonstrated, using rats in Pavlovian preparations, that many "deficits" in learned behavior (e.g., overshadowing, blocking, and effects of preexposure to conditioned or unconditioned stimuli [CSs or USs]) are due, at least in part, to processes that occur at the time of expression, rather than during acquisition as is assumed by most contemporary models of associative learning. Our data prompted us to reject the traditional associative position, which states that over a series of training trials subjects learn only a few summary statistics, and encourages the view that subjects encode a rich, nearly veridical representation of their experiences. To direct our research, we developed the "comparator hypothesis," which is a performance-focused rule for translating knowledge acquired through simple spatio-temporal contiguity into behavior. As we and others published ever more data problematic for the traditional associative emphasis on acquisition processes, researchers started acknowledging the importance of post-training information processing. Some investigators concurred with our performance-focused account of these findings. But other researchers modified existing acquisition-focused models to allow new learning about a previously trained CS on post-training trials on which the CS is absent. These revised acquisition-focused models can account for much of the data that we had viewed as uniquely supportive of our performance-focused account. This stimulated us to extend the comparator hypothesis, formalize it in a mathematical implementation (SOCR), and design experiments for which these new acquisition-focused models and the extended comparator hypothesis make divergent predictions. The proposed experiments will discriminate between these two types of models, thereby contributing to a more complete understanding of learning, memory, and acquired behavior, and directing the agenda of scientists concerned with the neurophysiological basis of learning and memory.
|
1 |