“Like déjà vu all over again”: Patterns of perseveration in two people with jargon aphasia

Abstract

Background: It has been argued that perseveration type corresponds to the level of breakdown, so that total perseveration (repetition of a whole word) involves reactivation of a previous word at the lexical level when the target word is insufficiently activated. A blended perseveration (repetition of part of a previous response) results from a failure of target activation at the phoneme level (e.g., Martin & Dell, 2007). This is challenged by the occurrence of nonword total perseverations, as these cannot be lexical retrievals (Hirsh, 1998). A further problem is the occurrence of long intervals between perseverations and their sources. Some authors have invoked semantic relationships to explain these intervals (e.g., Martin, Roach, Brecher, & Lowery, 1998).

Aims: This study examines the perseveration of two individuals with jargon aphasia and explores the proposal that while some perseveration may result from reactivation of recent responses as described above, others are built around default phonology, resulting in stereotypical errors.

Methods & Procedures: Tests of naming, reading, and repetition were administered. Responses were analysed to determine: the extent of perseveration; the occurrence of long intervals between perseverations and their sources; patterns of phoneme use; the occurrence of nonword total perseverations.

Outcomes & Results: Both individuals produced large numbers of perseverative responses. Perseverative responses following lengthy intervals could not be explained by semantic relationships. For each participant certain consonants were favoured and evidence was found of an interaction between the occurrence of perseveration and these favoured consonants. The possibility that word and nonword total perseverations arose from different sources was rejected because no difference was found in the use of the favoured phonemes in the two types.

Conclusions: The findings support the theory of two mechanisms for perseveration. The first is local, occurring when residual activation overrides incoming activation. This is confined to a single utterance and appears closely after the original occurrence. The second type is global, occurring across different contexts over time. It originates from default phonology when incoming activation is unavailable at the phoneme level. Both total and blended perseverations may result from this mechanism. Word total perseverations may be favoured because of feedback from the phoneme level to the lexical level. Ideas for future research and implications for intervention are discussed.

Keywords:

• Aphasia
• Jargon aphasia
• Perseveration
• Stereotypy
• Nonwords

Acknowledgments

We thank RS and TK and their families for their patience, participation, and support. We also thank Lyndsey Nickels for her help in accessing the CELEX database. Funding for this project was received from Bexhill and Rother (NHS) PCT.

Notes

*Attributed to Lawrence Peter “Yogi” Berra (1925–), baseball player for the New York Yankees.

¹There is a discrepancy between the numbers of non-perseverative errors in Tables 1 and 3. For example, in Table 1, 113 of 150 errors in RS's naming trials were perseverative, implying that 37 errors were non-perseverative. However, in Table 3, there are 29 non-perseverative errors in RS's naming. The reason for this is the removal of semantic errors of phonological distortions of semantic errors at the start of the phoneme frequency distribution analysis (Question 3a). This is a cautious treatment in this analysis: by removing semantically motivated errors from the non-perseverative errors we would expect to increase the proportions of stereotypical phonology, thus making the finding of a significantly higher rate of stereotypical phoneme occurrence in the perseverative than the non-perseverative errors less likely.

²We are very grateful to Hugh Buckingham for this suggestion in his review of the paper