Common mechanisms underlying perseverative and non‐perseverative sound and word substitutions

This research was supported by grants from the National Institutes of Health (National Institute of Deafness and Communication Disorders) R01 DC01924 to Temple University (PI: N. Martin) and R01 DC00191 to Moss Rehabilitation Research Institute (PI: M.F. Schwartz). We thank Paula Sobel and Adelyn Brecher who were responsible for patient testing, scoring, and data management. We are also grateful to Dana Bitetti, Lianne DiMarco, Stephanie Tempest, and Francine Kohen for their assistance in data organisation and analyses. Finally, we thank Myrna Schwartz for helpful discussions that contributed to the overall work.

Abstract

Background: Perseverations of sounds and words are common errors in aphasia. Understanding their mechanisms is of considerable interest to theories of word retrieval and also to treatment of anomia. Here, we explore the hypothesis that perseveration errors are generated by the same mechanisms as non‐perseverative errors: weak activation of the intended word in the context of a competition from other activated words.

This research was supported by grants from the National Institutes of Health (National Institute of Deafness and Communication Disorders) R01 DC01924 to Temple University (PI: N. Martin) and R01 DC00191 to Moss Rehabilitation Research Institute (PI: M.F. Schwartz). We thank Paula Sobel and Adelyn Brecher who were responsible for patient testing, scoring, and data management. We are also grateful to Dana Bitetti, Lianne DiMarco, Stephanie Tempest, and Francine Kohen for their assistance in data organisation and analyses. Finally, we thank Myrna Schwartz for helpful discussions that contributed to the overall work.

Aims: In analyses of perseverative and non‐perseverative naming errors of 94 individuals with aphasia, we aimed to show that the only difference between the two kinds of errors is that perseverated words and sounds have an increased probability of being retrieved instead of the target word because of their residual activation potential from their prior activation.

Methods and Procedures: Correlational analyses were conducted to test (1) an interactive activation model's (Schwartz, Dell, Martin, Sobel, & Gahl, 2006) account of the occurrence of sound and whole‐word perseverations, (2) distributions of perseverative and non‐perseverative across error categories, and (3) the relationship between weakened connection strength between semantic and lexical representations and the occurrence of “no response” errors.

Outcomes and Results: Our analyses indicate that whole‐word perseverative and non‐perseverative errors are associated with weak spreading of activation between semantics and the target word form, and that sound perseverations and non‐perseverations are associated with weak spreading of activation between the target word form and its corresponding sounds. Additionally, distributions of perseverative and non‐perseverative errors across error categories are strongly associated. Finally, the occurrence of “no response” type errors is associated with weak semantic activation and the occurrence of whole‐word perseverations.

Conclusions: The data from this study support a model of word and sound perseverations in which their occurrence is attributed to the same mechanisms as non‐perseverative errors: weakened activation of the intended word in the context of a competitive activation process in which other word forms that are related to the target word have potential to be retrieved instead of the target. Potential perseverated words and sounds participate in this process, but have an additional boost to their activation levels because of residual activation potential.

Notes

This research was supported by grants from the National Institutes of Health (National Institute of Deafness and Communication Disorders) R01 DC01924 to Temple University (PI: N. Martin) and R01 DC00191 to Moss Rehabilitation Research Institute (PI: M.F. Schwartz). We thank Paula Sobel and Adelyn Brecher who were responsible for patient testing, scoring, and data management. We are also grateful to Dana Bitetti, Lianne DiMarco, Stephanie Tempest, and Francine Kohen for their assistance in data organisation and analyses. Finally, we thank Myrna Schwartz for helpful discussions that contributed to the overall work.