Cognitive Neuroscience and Neuropsychology

neurofeport

WE have studied the distortion of perceived time in a
patient with left neglect. This patient consistently over-
estimated the duration of stimuli in the neglected space.
Overestimation was observed both with an interval
comparison (300/700 ms) and with a time production
(1s) paradigm. We suggest that encoding duration in
the hundreds of milliseconds range is a process based on
an internal clock mechanism. The functioning of that
clock varies as a function of the processing load.

Key Words: Time perception; Time production; Spatial
attention; Visuo-spatial neglect

NeuroReport 7, 2111-2114 (1996)

Time perception in a
neglected space

Gianpaolo Basso, Paolo Nichelli,“4
Francesca Frassinetti and
Giuseppe di Pellegrino

Dipartimento di Patologia
Neuropsicosensoriale, Sezione di Neurologia,
Universita di Modena, Via del Pozzo, 71, 41100
Modena, Italy

©ACorresponding Author

Introduction

Space and time are two fundamental matrices of
human behaviour. Studying patients with cerebral
lesions can help us to understand the physiological
basis of processing these two dimensions. Challeng-
ing observations that constrain our views of spatial
attention have been derived from patients affected
by visuo-spatial neglect, a disorder of orienting atten-
tion to stimuli contralateral to a cerebral lesion.
Patients with left-sided neglect may fail to orient
attention to the left side of a display even when
stimuli are presented in the right (intact) visual
field.'> These patients show reduced left side
processing despite normal visual function and consti-
tute a unique opportunity to test whether memory-
based models or internal clock models can account
for subjective duration. According to memory-based
models, the duration of an interval is extracted either
from the amount of information’ or from the number
of changes (cognitive, environmental, emotional, etc.)
occurring during the interval.3 Clock models main-
tain that subjective durations are a function of the
amount of attention allocated to the passage of time.

Decreased processing from the neglected side would
decrease the amount of information and the number
of changes extracted from that side but would make
more resources available to the internal clock. As a
consequence, memory-based and clock models would
respectively predict under- and over-estimation of
stimuli presented at a neglected location.

We have challenged these two classes of time
perception models by studying the distortion of
perceived time in a patient with left neglect. His
performance at different spatial locations was com-

© Rapid Science Publishers

pared with that of six matched normal controls (NC)
and one hemianopic patient without neglect (HP).

Materials, Methods and Results

A.M. is a 59-year-old right-handed man who was
examined 2 months after an ischaemic stroke, which
caused a left hemianopia and a moderate motor
impairment to the left limbs. A CT scan showed a
large lesion involving the right temporal-occipital
region, part of the thalamus, and the posterior
arm of the internal capsule. A.M. was alert and coop-
erative, with no general intellectual deficit. However,
he showed a severe left-sided visuo-spatial neglect.
On cancellation tasks (Bells’ Test* and Mesulam
Cancellation Test”) he consistently omitted crossing
out several items on the left side of the display. When
we asked him to bisect lines of different length, he
displaced the midline to the right of the true centre.
When asked to describe a room, both by direct
inspection or from memory, he omitted several items
on the left side.

In three experimental sessions we presented A.M.
and six normal controls with a visual display consti-
tuted by a matrix of 5 x 6 red circles (Fig. 1). Circles
were enclosed in a grey rectangle that appeared
always to the right of a fixation cross, in A.M.’s sound
hemifield. A black dot displayed inside a red circle
was the target stimulus. Subjects were sitting in front
of the computer screen. The grey rectangle of the
stimulus pattern subtended 8° x 9.5° and was placed
1° to the right of the fixation cross. The target stim-
ulus was a 0.2° black dot appearing in the centre of
a red circle. Fixation was monitored by an experi-
menter sitting at the rear of the computer screen.

Vol 7 No 13 2 September 1996 2111
neuro(peport

Experiment 1: Experiment 1 aimed to assess
whether A.M.’s visuo-spatial neglect was also present
within his right visual field. We asked A.M. and NC
subjects to verbally report dots appearing at random
for 200 ms, either on the right (first and second
column of the matrix) or on the left side (fourth or
fifth column) of the display. A total of 96 stimuli
were displayed, one at a time.

A.M. omitted 50% stimuli in the two leftmost
columns and 18.7% in the two rightmost columns
(x? = 10.39, d.f. 1, p = 0.001), while no NC subjects
failed to report any stimulus at any location. Based
on the presence of neglect in the right visual field,
we decided to compare A.M.’s subjective duration of
stimuli appearing in neglected and non-neglected
parts of the display, while still presenting stimuli in
the intact visual field.

Experiment 2: Subjects were examined in two sepa-
rate sessions, cach constituted by a training and a
testing phase. During training, 10 long (700 ms) and
10 brief (300 ms) stimuli were presented in one of
the two circles marked with an asterisk in Figure 1.
In each testing phase 160 stimuli (40 long and 40 brief
for each location) were presented in a pseudorandom
order in either the left or the right circle and subjects
were asked to discriminate between them.

No NC made any discrimination error in any
condition. In the training session, when stimuli were
repeatedly presented in the same location, A.M. made
no error. In the experimental session, when stimulus
position was randomly varied, A.M. misjudged
59.2% and 13.4% respectively of brief and long
stimuli presented at the left location (x? = 23.09, dif.
1, p < 0.0001). Conversely, he misidentified 8.6% of
brief and 45.2% of long stimuli appearing at the
right location (x? = 23.57, d.f. 1, p<0.0001). In
conclusion, his temporal perception was affected by
spatial position. He tended to judge left and right
stimuli respectively longer and shorter than they
were. A.M.’s pattern of responses can be interpreted
either as overestimation of stimuli on the left side of
the display or as underestimation of those on the
right side. Alternatively, his time perception could be
distorted in both fields. To distinguish between these
alternatives, we devised a procedure (Experiment 3)
that allowed us to compare A.M.’s subjective dura-
tions of left and right-sided stimuli with those of
control subjects. Moreover, to exclude that any
distortion of subjective duration was a consequence
of left hemianopia rather than of unilateral neglect,
we added an hemianopic patient (HP) as a further
control subject. HP is a 56-year-old male who had
undergone a right occipital infarction 2 years before
his testing. Computerized visual field examination
demostrated left homonimous hemianopia without

2112 Vol 7 No 13 2 September 1996

G. Basso et al

FIG. 1. Example of the stimulus pattern.

macular sparing. No other neurological deficits were
detected. Furthermore, at the task described in
Experiment 1, HP did not omit any stimulus at any
location.

Experiment 3: In the third experiment, trials
consisted of a variable wait (800-1200 ms) followed
by the target stimulus. The subject were required to
press a key to let the stimulus stay on the screen for
1s. During training phase, 96 stimuli were displayed
in a position 1° over the fixation cross and feedback
on timing accuracy was provided with a + 150 ms
tolerance interval (correct, too short, too long). In
two separate experimental sessions, a total of 480
stimuli (40 stimuli for each possible location)
appeared in a pseudorandom order in any circle of
the the first and fifth column. No feedback was given
at this time. Note that, since we define overestima-
tion as a ratio of > 1 between estimated (predefined)
and true (produced) duration, producing intervals
shorter than expected is indicative of temporal over-
estimation.

As shown in Figure 2, A.M. produced an interval
of 928 ms when the stimulus was displayed in the
leftmost column. This interval was significantly
shorter than the duration he was trained to produce
at the fixation center (t,,, = 3.369, p = 0.0009). This
finding indicates that he was overestimating stimuli
appearing on the left side of the display. Stimuli
appearing in the rightmost column remained on the
Time perception in a neglected space

neuro(peport

1300
1250 J
1200 |
1150 J
1100 1

1050 |

1000 | i
950

900 |

850

4 AM
-@ NC
HP

Temporal Production (msec)

dx sx

FIG. 2. Mean and standard error of the intervals produced by A.M.,
normal controls (NC) and an hemianopic patient (HP) in the left and
in the right part of the display.

screen for an average of 1003 ms, not significantly
longer than the trained duration (t,;, <1, n.s.). On
average, A.M. produced intervals 75 ms shorter in the
left than in the right side of the display (tyy = 2.493,
p=0.01), while neither HP nor any NC subject
produced significant duration differences in relation
to stimulus position (time differences ranged in NC
subjects between +35 and -10 ms). NC_ subjects
produced intervals of 1233 ms and 1242 ms, respec-
tively in the leftmost and in the rightmost columns.
HP produced 1078 ms at the left and 1081 at the right
location. In conclusion, A.M. overestimated stimuli
on the left side. NC subjects and HP showed a
consistent trend toward underestimating both left and
right-sided stimuli. However, no control subject
showed any estimation difference according to the
stimulus side.

Discussion

Taken together, the data from our three experi-
ments demonstrate a somewhat counterintuitive
spatial distortion of temporal perception. A.M.
perceived durations to be longer in that part of a
display where he neglected 200 ms stimuli. Since it is
conceivable that less information is available from
neglected locations, our data challenge the hypoth-
esis that duration is derived from the information
stored in memory during the interval.

To account for the pattern of temporal order and
simultaneous judgments in normal subjects, Stelmach
and Herdman* proposed a model in which the
temporal response of the visual system to a pulse of
light is a skewed bell curve. According to this model,
responses to impulses at unattended location would
produce a broader function, i.c. the visual response
would take more time to reach a maximum and more
time to subside. On the contrary, responses to stimuli
at attended location would be more brisk, with the
response function reaching a maximum more quickly

and subsiding earlier. This theory anticipated that
durations of unattended stimuli, due to their broader
temporal profile, would appear to be longer, whereas
duration of stimuli appearing at attended locations
would appear to be shorter. Our results support this
prediction.

As HP performed like NC subjects in the temporal
production task, A.M.’s distorted subjective duration
cannot be attributed to the presence of hemianopia
but rather to left-sided visuo-spatial neglect. i.e. to
defective allocation of attentional resources over the
visual display. We hypothesize that attentional allo-
cation affects the temporal profile of visual responses
via an internal clock mechanism.’ An_ internal
clock consists of an accumulator that sequentially
summates quantal time units produced by a pace-
maker. The result obtained by the accumulator must
then be compared. to values in a reference memory
system. In this framework, there are two possible
explanations for the local temporal overestimation we
have observed in A.M. According to one account, the
pacemaker rate is constant and independent from the
processing load, which only competes with the accu-
mulator’s reading of the pacemaker. Reduced visuo-
spatial processing from the neglected side may
increase resources available to the accumulator and
result in a greater number of pulses in the time unit
when stimuli are presented in the neglected field.
Alternatively, the pacemaker rate might be affected.
We may view pacemakers as reverberatory circuits
containing several neurons with an_ oscillating
activity.!° Some of these neurons have input and/or
output connections. We hypothesize that, under
resting conditions, the oscillation stabilizes at a
certain rate. Incoming information, by requiring
extra computation, could slow it down. As a conse-
quence of either mechanism, normal. subjects after
training with repeated presentation at the same visual
location (i.e. with minimal visuo-spatial process-
ing requirement), underestimated produced intervals
when, during testing, they were required to distribute
their attention over a larger visual display. Visuo-
spatial neglect, by decreasing processing load, could
either increase the number of pulses counted by the
accumulator or accelerate the oscillation rate, which
would ultimately cause temporal overestimation of
stimuli appearing at the neglected location.

Conclusion

This is the first report to show a link between a
reduction in cognitive resources (caused by visuo-
spatial neglect) and distortion of subjective stimulus
duration. We have demonstrated that visuo-spatial
neglect can cause overestimating stimulus duration
at a neglected location. This finding supports models

Vol 7 No 13 2 September 1996 2113
neurofeport

G. Basso et al

mantaining that short durations are computed via
internal clock mechenisms. We argue that the speed
of the clock is affected by concomitant processing.
Neglect, by reducing visuo-spatial processing, could
speed up the rate of the internal clock. This would
ultimately cause intervals to be judged longer than
when they are measured with a slower clock.

ACKNOWLEDGEMENTS. This research was supported by a grant from the Italian

National Research Council to P.N. We would like to thank Jordan Grafman for
his helpful suggestions and for his support.

Received 14 May 1996;
accepted 4 June 1996

References

Driver J and Halligan PW. Cognit Neuropsychol 8, 475-496 (1991).

Driver J, Baylis GC and Rafal R. Nature 360, 73-75 (1992).

Behrmann M and Moscovitch M. J Cogn Neurosci 6, 1-16 (1994).

Ornstein RE. On the Experience of Time, Middlesex, UK: Penguin, 1969.

Block RA. Contextual coding in memory. Studies of remembered duration.

In: Michon JA and Jackson JL, eds. Time, Mind, and Behavior. Berlin:

Springer-Verlag, 1985.

6. Gauthier L, Dehaut F and Joanette Y. J Clin Exp Neuropsychol 11, 49-54
(1989).

7. Mesulam MM. Attention, confusional states and neglect. in: Mesulam MM,
eds. Principles of Behavioral Neurology . Philadelphia, PA: Davis, 1985:

8. Stelmach LB and Herdman CM. J Exp Psych: Human Perc and Perfor 17,
539-550 (1991)

9. Gibbon J, Church RM and Meck WH. Ann NY Acad Sci 423, 52-77 (1984).

10. Miall RC. Neural Computation 11, 359-371 (1989)

Paeena

General Summary

We have examined subjective stimulus duration in a patient (A.M.) affected by left visuo-spatial neglect, a disorder that impairs the
ability to orient attention contralateral to a cerebral lesion. We have demonstrated that A.M. neglected left sided visual stimuli even
when they appeared in the right hemifield. At an interval comparison task, he judged stimuli on the left to be longer than those on
the right side of the display. Temporal overestimation in the neglected field was confirmed with an interval production pardigm.
These findings support the view that stimulus duration in the hundreds of milliseconds range is computed via an internal clock
mechanism. The speed of the clock could be affected by concomitant processing. Reduced processing following visuo-spatial neglect

could speed up the clock rate and cause temporal overestimation.

2114 Vol 7 No 13 2 September 1996