384
R. D. Oades et al.
Few studies have focused on the relation of symptoms to attentional performance.
Comparisons between studies are dangerous considering the different requirements for
performance and illustrate the ease of drawing incorrect inferences. Thus, thought disorder
has been reported to correlate with distractibility on the digit span task (Oltmanns et al.,
1978). However, in the CB task irrelevant stimuli did not catch the attention of thought-
disordered patients; thought disorder correlated with normal CB. However, the tasks were
differently set up: in the current CB task the stimuli were spatially separated and there was
no task requirement to suppress response to the added stimulus.
In another example, Green & Walker (1986) noted that positive symptoms and low signal
detection (d-prime) related to poor digit-span performance. In the present study those with
positive symptoms performed CB better and had better d-prime scores than NP patients
(unpublished). Indeed, as CB attenuation reflects learning about the stimulus in question,
poor signal detection is hardly a viable explanation of NP performance.
Further, digit-span scores did not relate to poor word association performance (Allen,
1990) which required “contextual processing". This feature may be more relevant to learning
about associations in the CB task than attentional span and relate to our proposal (above)
on the role of perseverative learning strategies. Lubow & Gewirtz (1995) have already
posited a critical role for “context” as an “occasion setter” for eliciting learned inattention.
They suggested that decreased latent inhibition in psychotic patients reflects an inability to
fully use this aspect of the context or to switch from controlled to automatic processing of
inconsequential events. It would be interesting to see if neuropsychological tests of frontal
lobe performance, where dysfunction is implicated in poor word fluency and perseverative
strategies (discussion in Elliott et al., 1995), would correlate differently with PH and
NP patients’ CB scores. Many studies find frontal dysfunction associated with negative
symptoms (for a review see Straube & Oades, 1992) and impaired CB has been reported in
frontally damaged animals (Oades et al., 1987).
The paranoid~non∙paranoid distinction
We have emphasized in the previous sections the associations with symptom-cluster.
This is important for two reasons. First, our subgroup distinction was based on the
presence∕absence of PH symptom-clusters defined by the SANS∕SAPS interview. Secondly,
current views on three types of schizophrenia explicitly refer to symptom-groupings and
not syndromes (Liddle, 1987; Liddle et al., 1992). Thus, earlier neuropsychological studies
of the performance of paranoid or non-paranoid patients were not necessarily based on the
same distinction used here.
The potential usefulness of differentiating paranoid patients was recognized long ago
(Kahlbaum, 1863), but the experimental study of paranoid vs non-paranoid cognitive style
only really got under way IOO years later (Silverman, 1964; Magaro, 1980). These authors
maintained that non-paranoid patients relied more on perceptual data and less on con-
notative or contextual aspects of situations to solve tasks. While more recent discussions
find this view contentious (e.g. Strauss, 1993), the idea is replaced with the suggestion that
positive symptoms reflect, in part, failures of automatic processing. To us this is not
inconsistent with earlier formulations. Paranoid patients may be able to cope with a