Taking the thumbs off the multidimensional scales in the debate on semantic memory and Alzheimer's disease: A comment on Storms et al. (2003).

Abstract

Verstraeten, and P. P. De Deyn (2003) bring into focus at least 3 important issues. First, the performance of even simple tasks may be affected by several basic cognitive processes. Second, the statistical variance that these processes contribute to task performance may be different for a group of participants with neuropsychological deficits than for normal control particpants. Third, claims regarding the nature of semantic memory in Alzheimer's disease (AD) reflect indeterminacy about the nature and assessment of semantic memory itself. Critically, their analysis suggests that the T. Shallice and E. K. Warrington (1979) criteria widely used to operationalize the semantic memory deficit in AD be carefully scrutinized and reevaluated.

What happens to semantic memory when formal thought disorder remits? Revisiting a case study

“…(See Figure 1.) Research into the neuropsychology of semantic memory has used a model proposed by Warrington and Shallice (1979), which suggests that there are two patterns of semantic memory impairment: first, degraded store type disorder (knowledge is lost), and second, impaired accesses type disorder (knowledge exists but the retrieval process is compromised). The validity of this distinction has been occasionally disputed (e.g., Milberg & McGlinchey, 2003;Rapp & Caramazza, 1993); however, these two patterns of semantic memory impairment in neurological disorders have achieved wide acceptance in the neuropsychological literature (see Storms, Dirikx, Saerens, Verstraeten, & De Deyn, 2003a, 2003b for discussions see Al-Uzri, in press; Al-Uzri, Laws, & Mortimer, 2004). Laws et al, 1999).…”

On what we cannot learn from proximity data.

“…Takane (2003) adds to this that no data-analytic method can overcome this lack of relevant information in the data. Milberg and McGlinchey (2003) come basically to the same conclusion by acknowledging that several elementary cognitive processes contribute in gathering the similarity data and that it is not even clear whether these different subprocesses account for the same amount of variance in the patient and in the control groups.…”

“…Taking into account both possible sorts of access problems, it would seem apparent that patients with an intact semantic storage perform worse in all three types of tasks. Details on the difficulties of distinguishing between access and storage problems are addressed by Hutchison and Balota, where this discussion is phrased in term of structure versus processing, and by Milberg and McGlinchey (2003), where the issue is put in a historical perspective.…”

“…Two of the commentaries focused on the Warrington and Shallice criteria. First, Milberg and McGlinchey (2003) argue that the criteria deserve further scrutiny and discussion, because these criteria were first formulated to describe an aphasia patient who was assumed to have a semantic storage deficit, whereas later research casts doubt on that assumption. They further question the performance/knowledge equivalence that underlies the criteria for semantic degradation.…”

Problems of reliability and validity with similarity derived from category fluency

“…More often than not the application of the procedure to category fluency data from two distinct groups has produced similarities that differ between the groups. However, there is debate about the origin of these differences and the inferences they warrant (Chan and Ho, 2003;Elvevåg and Storms, 2003;Hutchison and Balota, 2003;Jarrold, 2003;Milberg and McGlinchey, 2003;Ober and Shenaut, 2003;Rogers, 2003;Storms et al, 2003aStorms et al, , 2003bTakane, 2003;Voorspoels et al, 2014). A prime objection to the method relates to a potential lack of reliability of the derived similarities, both for patient and control groups, which might make one erroneously conclude that the semantic structures of two groups differ.…”