Acta Psychologica 114 (2003) 41–65

www.elsevier.com/locate/actpsy

Spatio-temporal working-memory
and short-term object-location tasks
use different memory mechanisms
Hubert D. Zimmer *, Harry R. Speiser, Beate Seidler
Department of Psychology, Saarland University, P.O. Box 151150, Saarbr€ucken D-66041, Germany
Received 1 February 2002; received in revised form 14 April 2003; accepted 8 May 2003

Abstract

Spatial short-term memory for objectsÕ locations was investigated in a spatial relocation
task. During maintenance, dynamic visual noise or spatial tapping were administered as visual
or spatial secondary tasks, respectively. Because memory for location should tap the visual
component of working memory, a visual but not a spatial secondary task should impair loca-
tion memory. In fact, neither of the tasks impaired memory (Experiment 1), although the
expected dissociation between visual and spatial components was clearly confirmed for a spa-
tio-temporal main task (Corsi test) (Experiment 2). We then contrasted location memory for
pictures of objects and of nonsense figures under visual interference. Real objects were relo-
cated much better than nonsense figures, and visual noise was again ineffective (Experiment
3). When spatial tapping was combined with the same material (Experiment 3a), again no in-
fluence on memory for locations of objects was observed and only a small influence on remem-
bering nonsense figures. We suggest that the Corsi and the relocation VSWM-tasks use
different memory mechanisms. The configuration of objects is reconstructed from perceptual
records in an episodic buffer, provided by the same structures that enable visual memory after
longer intervals. Rehearsal is not necessary for the persistence of these traces. In contrast, in
the Corsi task remembering, a temporal sequence across homogeneous locations needs spatio-
temporal marking and therefore active rehearsal of the locations by shifting spatial attention.
A spatially demanding secondary task during retention interrupts this rehearsal.

2003 Elsevier B.V. All rights reserved.

PsycINFO classification: 2343

Keywords: Spatial learning; Visual interference; Corsi task; Irrelevant pictures; Spatial attention

*
Corresponding author.
E-mail address: huzimmer@mx.uni-saarland.de (H.D. Zimmer).

0001-6918/$ - see front matter
2003 Elsevier B.V. All rights reserved.
doi:10.1016/S0001-6918(03)00049-0
42 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65

1. Introduction

Data from behavioral (cf. Logie, 1995) and neuropsychological experiments (cf.
Smith & Jonides, 1997) support the assumption that humans have a specific subsys-
tem called visuo-spatial working memory (VSWM) that is specialized for the pro-
cessing and temporary storage of visual and spatial information (Logie, 1995).
Although the specific components and the details of processes within this module
are still under discussion, there is agreement about a number of principles and con-
straints. The VSWM can store visual as well as spatial information, and it has a spe-
cific rehearsal process (e.g., Logie, 1986); the amount of to-be-retained information
is limited (e.g., Kemps, 1999; Logie, Zucco, & Baddeley, 1990), and it is prone to in-
terference if a secondary task has to be performed (e.g., Logie & Marchetti, 1991;
Quinn & McConnell, 1996a).
Though these assumptions are well supported, other aspects of this module are
not clear. A central issue among these topics is the question of the content to which
the VSWM is dedicated. Sometimes it is supposed to be a device for processing and
storage of any visual information, be it spatial or not, sometimes one for storage of
any spatial information independent of the type of sensory input, and sometimes it is
argued that only spatio-temporal information is stored in the VSWM. The topic of
the present paper is memory for a specific subset of spatial information. We want
to investigate short-term memory for spatial locations of objects simultaneously pre-
sented, and we explore to what extent a visual or spatial component within the
VSWM contributes to this memory performance. For that purpose we contrast
objectsÕ location memory with and without a visual and spatial secondary task,
respectively, and we compare the results with secondary task effects known from spa-
tio-temporal tasks.
Earlier, when the multiple component model of working memory was proposed
(Baddeley & Hitch, 1974), the visual spatial scratch pad (VSSP) was distinguished
from the verbal component. The VSSP was soon considered as a spatial processing
device that is not specifically related to vision (Baddeley & Lieberman, 1980). Later,
from the work of Logie (1986), it was accepted that visual information is also pro-
cessed within the VSSP. In line with this position, Gathercole and Baddeley (1993)
argued that the VSSP represents visual and spatial information and therefore also
represents the distribution of objects in space. In the following period, a distinction
between object memory and proper location memory was proposed. Postma and De
Haan (1996) suggested that the spatial components, i.e. the coordinates of the occu-
pied spatial positions, are stored within the VSSP, whereas object memory (identity)
is enabled by other components (e.g., the verbal one). In contrast, Logie (1995) sug-
gested that within the VSWM a visual and a spatial sub-component must be distin-
guished. The selective interference between visual (object) information and visual
secondary task, and spatial (location) information and spatial secondary task were
quoted as support for this assumption (e.g. Logie & Marchetti, 1991; Tresch, Sinna-
mon, & Seamon, 1993).
Sometimes, these two components were attributed to two different percep-
tual pathways: The parvocellular system supposedly processing and storing visual
 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65 43

information (color or shape) and the magnocellular system supposedly processing

and storing spatial information (Hecker & Mapperson, 1997; Kessels, Postma, &
de Haan, 1999). However, how this distinction is important for memory is controver-
sial (Goodale & Humphrey, 1998). It may be that the two systems do not directly
contribute to the storage of information, but that they only modulate the quality
of input to a supra-modal working memory.
Related to this distinction are results from recent neuropsychological experiments
strongly supporting the assumption that object information and spatial information
are processed by different neural structures (e.g., Mecklinger, 1999; Mecklinger &
Pfeifer, 1996; Smith et al., 1995). In these studies too, static spatial layouts were used.
Based on their results the authors concluded that different brain structures perform
storage of object content and of spatial configuration. These structures, however, are
interconnected and this net constitutes visuo-spatial short-term memory. If this is
valid, working memory is no longer regarded as a static unit, but as a flexible collec-
tion of components that are assembled according to the subjectsÕ task. The specific
VSWM component suggested by the multiple-component model would thus be rep-
resented by a specific collection of object and spatial processing modules.
However, it is still an open question which information should be considered as
visual and which one as spatial in short-term memory. Sometimes static configura-
tions of spatially distributed objects are called spatial, whereas at other times the
use of the term ÔspatialÕ is seemingly confined to conditions in which ÔobjectsÕ change
their locations over time. Logie distinguished between a Ôvisual cacheÕ and a spatial
working memory proper. The spatial working memory ‘‘retains dynamic information
about movement and movement sequences, and it is linked with the control of phys-
ical actions’’ (Logie, 1995, p. 2). Maintenance of this spatial information needs a spa-
tial rehearsal process. Support for this distinction is seen in the interference of
spatially guided movements with memory for sequences of locations (the Corsi block
test) (e.g., Logie & Marchetti, 1991; Quinn & Ralston, 1986; Smyth, Pearson, &
Pendleton, 1988; Smyth & Scholey, 1994). According to this definition, spatial would
only refer to dynamic spatial information, i.e. either a temporal sequence of locations,
trajectories of objects moving in space, or a body movement in space.
In contrast, the static information about spatial locations of items as provided by
configurations is considered visual. ‘‘A visual representation in working memory
might involve retention of static visual arrays which incorporate geometric proper-
ties of the layout of objects. . .’’ (Logie, 1995, p. 78). This information is supposed
to be held within a visual cache. As a consequence, studies in which patterns of filled
and empty cells are to be remembered are interpreted as studies on visual and not
spatial working memory (e.g., Logie & Pearson, 1997). According to this view, re-
membering configurations of objects is a visual task which is performed by the visual
cache and it is not a proper spatial task that is processed within the spatial compo-
nent in the VSWM.
This view, however, is at odds with studies in which arrays of objects were used to
investigate spatial short-term memory. In these experiments, objects were presented
scattered on an otherwise empty screen, and during testing participants had to either
reconstruct objectsÕ positions into an empty matrix or they had to recognize whether
44 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65

the objects were presented at the same locations as during study (e.g., Mecklinger,
1999; Mecklinger & M€ uller, 1996; Postma & De Haan, 1996; Schumann-Hengsteler,
1995). 1 In this research tradition, configurations of items are common stimuli that
are typically used to investigate spatial information within the visuo-spatial short-
term memory. In contrast to the VSWM approach, in this view remembering the con-
figuration of objects is considered a spatial task. Dynamic information is not an issue
in this context, but spatial processing is contrasted with object processing concerned
with identity of the items.
Finally, in a third research tradition configurations of objects are also considered
as spatial information. Even more, it is assumed that locations are obligatory fea-
tures of visual short-term memory that are automatically encoded. In these studies,
indirect effects of configurations on the processing of visual non-spatial information
were investigated. The subjectsÕ task was to match the content of two stimuli in an
S1–S2 paradigm with a retention interval of a few seconds. In these studies it was
shown that item recognition (Santa, 1977; Zimmer, 1998) and even color recognition
(Jiang, Olson, & Chun, 2000) were impaired if the configuration of items had chan-
ged from study to test, although the configuration itself was irrelevant for the deci-
sion. This speaks in favor of the assumption that the configuration of objects is
automatically processed and stored in memory, although only visual information,
e.g. form or color, was relevant for the task. From this point of view, visual and spa-
tial information may be integrated within visual-short-term memory and would not
run separately as it is suggested by the distinction of a visual and a spatial working
memory component. This may be the case even though object and spatial informa-
tion are represented in different brain structures. If both components are bound, they
appear as units during processing, presupposed that both elements are represented
within the same memory record.
Obviously, conflicting assumptions exist in respect of the functional role of spatial
information provided by the locations of objects. On the one hand, within the
VSWM model, locations of objects are interpreted as a visual component that is sep-
arate from the dynamic component and only the latter is considered as spatial. On
the other hand, outside the VSWM tradition, locations of objects are regarded as
a typical spatial representation that is distinguishable from visual representations
of, for instance, shape and color. This variability can even be found in the work
of Logie himself. Logie (1995, p. 92) discussed memory for the spatial location of
a dot on a blank screen (Tresch et al., 1993) in the context of spatial memory,
whereas he discussed memory for the locations of circles on a blank screen (Morris,
1987) as visual (p. 90).
Because of these controversial positions on processing of location information in
a VSWM task, we wanted to investigate memory for the location of objects in a sec-
ondary task paradigm, in order to decide whether memory for locations of objects is
a spatial or a visual task. The logic behind this paradigm is that performance based

1
Similar experiments were conducted to investigate spatial long-term memory (e.g., Chalfonte &
Johnson, 1996; Naveh-Benjamin, 1987; Pezdek, Roman, & Sobolik, 1986).
 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65 45

on a specific working memory component is impaired if the secondary task uses the
same component as the main task. For our purpose, the important point is that two
secondary tasks are at hand which are considered exclusively visual or spatial by pro-
ponents of the VSWM model. Spatial tapping is regarded as a spatial secondary task,
whereas dynamic-visual-noise is regarded as a visual secondary task. Therefore, if
memory for the location of objects relies on the visual component, visual noise
should impair memory, but not spatial tapping. On the contrary, if location memory
is a spatial task, interference should be reversed. We investigated short-term memory
for locations of objects in a free relocation task. We presented four objects on arbi-
trarily selected spatial positions on an otherwise empty screen. The configuration of
the objects was to be reconstructed by our participants after a brief retention inter-
val. In the control condition the retention interval was empty, and in the experimen-
tal conditions it was filled with one of the secondary tasks.
We selected the two secondary tasks according to the general agreement on the
type of induced interference. In (a) the spatial tapping task, subjects tap a series
of buttons arranged in an array at a constant speed, using their forefinger. This task
was shown to interfere with dynamic spatial memory in many experiments (e.g., Lo-
gie & Marchetti, 1991; Morris, 1987; Quinn & Ralston, 1986; Smyth et al., 1988;
Smyth & Scholey, 1994). Because Logie and Marchetti (1991) did not find any inter-
ference when they used this task together with a visual main task (color memory),
they considered spatial tapping a purely spatial secondary task. In contrast, dynamic
visual noise is considered a visual secondary task. In this task participants view a
black and white matrix with some filled cells, and a part of these cells randomly
switch their color every 100 ms. Such a task was successfully used by Quinn and
McConnell to interfere with visual imagery main tasks (loci and pegword mnemon-
ics) (e.g., Quinn & McConnell, 1996a, 1996b, 1999). The authors concluded from
their results that dynamic visual noise interferes selectively with the visual cache,
i.e. the visual component of the VSWM (Quinn & McConnell, 1999).
All secondary tasks were applied only during the retention interval. Hence, we in-
terfered with maintenance and not with encoding or retrieval of information. We did
this because we were interested in the temporary storage of object locations and not
in the processes of encoding or retrieval of spatial information. We also wanted to
keep the product of spatial processing intact so that the quality of the memory trace
was not impaired by the secondary task. Most interference experiments on working
memory have used a concurrent secondary task, but in some of them secondary tasks
were presented only during retention just as we planned to do. In these experiments
interference effects were observed for sequences of locations and for colors (Logie &
Marchetti, 1991), in the Corsi task (Smyth & Scholey, 1994), the Brooks task (Toms,
Morris, & Foley, 1994) and finally, for positions of dots and for the shape of geomet-
rical objects (Tresch et al., 1993). An exception is marked by the study of Morris
(1987) who observed only weak and non-significant interference effects on memory
of circlesÕ locations. The most extreme position was held by Phillips (1983), who
speculated that visual short-term memory can store only one pattern––usually the
one most recently perceived––and any additional new information would displace
the formerly encoded pattern from short-term memory. Given these results, either
46 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65

the visual noise or the spatial tapping task should show an interference effect, de-
pending on whether short-term memory for spatial locations is using the visual or
the spatial component of VSWM, respectively.

2. Experiment 1

In this experiment participants performed a spatial relocation main task. Partici-

pants had to reconstruct the configuration of four objects after a short retention in-
terval. We used a set size of four objects because it was suggested that this was roughly
the maximum capacity of VSWM (Luck & Vogel, 1997). In the critical conditions a
secondary task was applied during the interval (tapping or presentation of visual
noise). Because visual noise can have different impact on processing, two types of vi-
sual noise were used. It was either static, i.e. a static randomly filled checkerboard was
presented and the pattern of black and white cells did not change, or it was dynamic,
i.e. the checkerboard pattern changed every 100 ms. The latter condition had a dy-
namic component, whereas the first did not. We used both visual secondary tasks be-
cause the dynamic task should have a stronger effect than the static one (Quinn &
McConnell, 1996a, 1996b). However, if none of these visual noise conditions had
an effect, we can be quite sure that we did our best to set up a visual interference.

2.1. Method

2.1.1. Participants
Eighty participants took part in Experiment 1. Thirty-two (one group with and
one without tapping) in Experiment 1a, and 48 in Experiment 1b (one group without
interference, one with dynamic noise, and one with static noise). Hence, all experi-
mental variations were manipulated between subjects. The participants were paid
for their participation. All were students from Saarland University, and about
two-third were female.

2.1.2. Material and design

For the main task we selected 64 multicolored objects from the Corel Mega Gal-
lery. These figures were vector graphics depicting everyday objects. Examples can be
found in the top half of Fig. 1. The pictures were sized so that they fitted into a
square of 50 · 50 pixels. When displayed, they were about 8 · 8 cm in size. With these
pictures, we constructed 16 different layouts, each consisting of four objects. The
objectsÕ positions were selected in a way that the configuration could not easily
be named, i.e. no global shape resulted for which names are available, e.g. a
square or an L-shaped figure. By doing so the possibility of verbal recoding was
minimized. 2 The locations were selected within an imaginary 9 · 9 matrix which,

2
We did not use an additional articulatory suppression task as we did not want to overstrain our
subjects who already had to cope with a complicated arrangement of main and secondary tasks.
 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65 47

Fig. 1. An illustration of the type of stimulus material (objects and artificial figures) used in the relocation
task. The configuration is not drawn to scale. Spatial distances between objects were larger in the used
configurations compared to objectsÕ size.

however, could not be inferred from the display. The boundary box of any object
was the same size as a matrix cell. Each subject processed 16 layouts. An example
of a layout is given in Fig. 1.
48 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65

A response board was constructed for the tapping task. Twelve response keys
were mounted in a 4 · 3 matrix on a separate pad with a distance of about 4 cm be-
tween rows and columns. The keys were micro switches which could be easily pushed
using the forefinger.
For the visual noise task we constructed a checker board of 30 · 30 cells. This
board was 450 · 450 pixels like the imaginary matrix so that it completely covered
the screen area on which the objects were presented. Half of the 900 cells were ran-
domly set to black. The remaining cells were white. The individual cells had no vis-
ible outlines so that the neighboring cells of the same color formed small block
figures. In the static visual noise condition, one such matrix was presented in
throughout each retention interval. In the dynamic visual noise condition, we ani-
mated this pattern. The complete matrix was subdivided in four fields of 15 · 15 cells
(quadrants), and within each of these quadrants 15 randomly selected cells changed
their color every 100 ms. The color was changed from black to white and vice versa.
This resulted in a changing rate of 600 cells per second.

2.1.3. Procedure
The whole experiment was run on a computer controlled by custom-made soft-
ware. The layouts were rear projected on a screen by an LCD projector into an
empty field of about 72 · 72 cm (450 · 450 pixels). This target field was marked by
a black frame. The viewing distance was about 1.5 m. The timing of a trial was as
follows: The four objects were simultaneously presented for 8 s, and a retention in-
terval of 10 s followed.
In the interference condition the task was administered during the retention inter-
val. With tapping interference, participants placed their hand on the response board
at the beginning of a trial and had to tap the keys during the retention interval,
which was indicated by the empty screen, at a rate of one key press per second. Tap-
ping was practiced at the beginning of the experiment. The keypad was placed beside
the subject and out of sight. Subjects had to start in the upper left corner, tap the first
column, and then continue with the next column. The key presses were registered by
a computer. When the test phase started, participants stopped their movement and
started to relocate the objects. In the visual noise condition, during the retention in-
terval, the screen showed the visual noise pattern, either static or dynamic. Subjects
were instructed to look at the screen and not to close their eyes. This was not checked
by any technical equipment, but the experimenter was present throughout the exper-
iment and due to the large size of the projection field it could easily be established
that subjects did watch the screen.
A free relocation test followed afterwards. During the test, the frame was shown
and the old objects were lined up at the left and right side of the field (two objects on
the left, two objects on the right). Participants moved the mouse cursor to an object,
picked it up by clicking the left mouse button, and moved it to its target position per
drag and drop. This way, the subjects had to process all four objects to reconstruct
the layout. Participants were allowed to reposition objects previously placed in a dif-
ferent position. When the subjects signaled that they had finished the relocation task,
the screen was cleared and the next trial started after an interstimulus interval of 1 s.
 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65 49

2.1.4. Results and discussion

We analyzed the positions to which the objects were relocated. The center of the
boundary box around a picture was taken as the remembered objectÕs position. To
estimate memory performances, we categorized the locations as either right or wrong
in order to obtain accuracy data. A location was counted as correct if the remem-
bered position was located within a circle with a diameter of 8 cm around the study
location (also the center of the object). This was the acceptance zone for correct po-
sitioning. In a further analysis, we calculated the exact metric distances between the
original and the remembered position. The two analyses based on proportion correct
and on exact metric distances yielded comparable results. To make it easier to com-
pare the relocation performances with the results of the following experiments, we
are only reporting the proportion correct. Fig. 2 depicts the proportion correct, de-
pendent on the type of interference.
A look at this figure immediately reveals that memory for objectsÕ locations was
not impaired by any of the secondary tasks. In order to test these differences we cal-
culated a one way analysis of variance with the factor type of secondary task (none,
spatial tapping, visual static noise, and visual dynamic noise). The variations be-
tween the conditions were small and not significant, F ð3; 76Þ ¼ 2:05,
MSE ¼ 0:014, p 6 0:12. A separate analysis for the tapping condition yielded no sig-
nificant interference effect, F ð1; 30Þ ¼ 1:32, MSE ¼ 0:014, p 6 0:26, and the same
was true for visual interference, F ð2; 45Þ ¼ 1:96, MSE ¼ 0:014, p 6 0:15. Numeri-
cally, performances in the interference condition were even better than in the baseline

1.0

0.9

0.8
Proportion Correct

0.7

0.6

0.5

0.4

0.3
Without With Without Static Dynamic
Spatial Tapping Visual Noise

Type of Secondary Task

Fig. 2. Relocation performances in the objectsÕ location task dependent on the type of visual secondary
task (left panel: tapping, right panel: visual noise). Note: the whiskers correspond to the 95% confidence
intervals of the mean.
50 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65

condition. An analysis of the performances in the secondary task (tapping) demon-

strated that subjects had a constant and a bit too fast tapping rate with an average
inter-tapping time of 832 ms, SE ¼ 11.
Obviously maintenance of memory for objectsÕ locations was not impaired––nei-
ther by a spatial nor by a visual secondary task. The confidence intervals of the
means make clear that this null effect is probably not due to insufficient power of
the statistical tests. We would therefore like to conclude for the moment that loca-
tion memory is not interrupted by additional input or active movements during re-
tention.

3. Experiment 2

This experiment served as a control condition for Experiment 1. In the light of

neuropsychological data we thought memory for the location of objects was more
likely to be a spatial than a visual task. We therefore expected a spatial, rather than
a visual interference. As suspected, no visual interference could be found. Perfor-
mances were numerically even a bit higher with than without a secondary task. How-
ever, contrary to our expectation, there was no spatial interference, either. Although
a small impairment of memory was found, it was far from being significant. This re-
sult does not support the idea of a resource limited spatial component. However, we
would feel more comfortable if this small difference in Experiment 1 could be consid-
ered as negligible compared to the spatial interference effect in a standard working
memory paradigm. To get this baseline we designed a Corsi task, which is generally
accepted as a spatial task according to the VSWM model (e.g., Fischer, 2000; Smyth
& Scholey, 1994). This task was run within the same visual field and using the same
locations as in Experiment 1. Participants had to remember a series of locations.
Each location was temporarily marked within a larger set of locations all holding
identical objects. After a brief delay, the sequence had to be reproduced by selecting
the correct locations in the correct temporal order.
We combined the Corsi task with the same spatial tapping or visual noise tasks we
had used in Experiment 1 as secondary tasks. If the secondary tasks cause the ex-
pected effects, performances in the Corsi block test should be impaired by tapping
(e.g., Smyth & Scholey, 1994), whereas perceiving visual noise––a task considered
as being purely visual (Quinn & McConnell, 1996a)––should not interfere with the
spatial main task.

3.1. Method

3.1.1. Participants
Eighty students from Saarland University participated in Experiment 2. As be-
fore, about two third of the participants were female. All participants were paid
for their collaboration. Again, we randomly assigned the subjects to five groups of
16 subjects, so that the type of secondary task was between subjects factor.
 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65 51

3.1.2. Material and design

The design of this study was the same as in Experiment 1. Only the main task had
changed. For the Corsi test, we depicted 16 circles evenly spaced in a 4 · 4 matrix
within a frame of 72 · 72 cm. All circles were identical and all were visible through-
out the study and test phase. During study, four of these circles were sequentially
marked as to-be-remembered locations. After a brief retention interval, participants
were required to reproduce the series of indexed locations on the screen. Subjects
studied 16 sequences of this kind in each condition.

3.1.3. Procedure
The experiment was run on a computer. During the study phase, participants saw
all 16 circles (outlines) rear projected by an LCD projector. On this array, a sequence
of four locations was presented. The four positions were marked at a rate of one per
second. In order to mark a position, we colored the inside area of the circle at the
target position blue. The color was visible for 500 ms, then it was reset to default
for 500 ms and then the next circle was colored. This procedure caused the impres-
sion of a dot that jumped from position to position. A retention interval of 10 s fol-
lowed. In the no-interference and tapping conditions, the screen was blank during
this interval. In the visual noise condition the noise was presented in this period.
Thereafter, the participants were tested. All 16 circles were shown again, and subjects
were required to indicate the correct locations in the correct temporal sequence using
the computer mouse. When a circle was marked, it flashed blue for 500 ms.

3.1.4. Results and discussion

We calculated the number of correct locations in terms of spatial and temporal
information. If, for example, location seven was marked at the second serial position
during study and this location was selected as third item during test, subjects received
no credit for this. This index takes into account that the Corsi task is a spatio-tem-
poral task. The average performances, dependent on the secondary task, are depicted
in Fig. 3(left panel and right panel).
It turned out that spatial tapping task strongly interfered with the Corsi test. Per-
formances were much better without (0.86) than with tapping (0.42). In contrast,
with the static visual pattern (0.83) and the dynamic visual noise (0.85), memory per-
formances were as high as without the secondary task (0.86). Tapping was accurate
but a bit faster than the instructed one beat per second. The mean interval in the tap-
ping task was 874 ms, SE ¼ 9, which is comparable to Experiment 1.
In a one-way analysis of variance with the factor type of secondary task (no spa-
tial tapping, spatial tapping, visual static noise and visual dynamic noise), we yielded
a highly significant effect, F ð3; 76Þ ¼ 52:71, MSE ¼ 0:016, p < 0:001. With tapping,
a strong interference effect was observed, F ð1; 30Þ ¼ 220:69, MSE ¼ 0:014,
p < 0:001, whereas in the visual noise conditions, no interference effect occurred,
F ð2; 45Þ ¼ 0:21.
Obviously, performances in the Corsi task, which is considered a spatial task, were
strongly impaired by spatial tapping (cf. also, Smyth & Scholey, 1994). Hence, spa-
tial tapping during retention is able to impair maintenance of visual spatio-temporal
52 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65

1.0

0.9

0.8
Proportion Correct

0.7

0.6

0.5

0.4

0.3
Without With Without Static Dynamic
Spatial Tapping Visual Noise

Type of Secondary Task

Fig. 3. Performances in the Corsi task dependent on the type of secondary task (left panel: tapping, right
panel: visual noise). Note: the whiskers correspond to the 95% confidence intervals of the mean.

information. Visual noise does not have this capacity, although the dynamic noise
has a high changing rate and a dynamic component. This corroborates the fact that
a spatial main task reveals the expected dissociation between the visual and spatial
secondary task.
For the sake of comparison with Experiment 1, we also calculated a second score.
In the Corsi task subjects had to reproduce a spatio-temporal sequence. The second-
ary task, in principle, could have selectively impaired temporal information, leaving
spatial information unaffected. In order to test this possibility we counted the num-
ber of correct locations independent of their temporal correctness. Hence, in this
analysis, we did not consider temporal order information as in the first analysis.
However, with this score, the results were the same as with the standard Corsi anal-
ysis. Performances were 0.89 without versus 0.51 with tapping; for the static-, dy-
namic-, and no visual noise conditions the performances were 0.88, 0.89 and 0.90,
respectively. This result demonstrates that the selective interference in Experiment
2 was not due to purely temporal errors, i.e. correct spatial positions but solely
wrong temporal order. Therefore, we conclude that memory for locations of the kind
used in the Corsi task is impaired by spatial tapping.
Obviously, spatial tapping has enough impact on spatial information processing
to induce an interference effect in a spatial main task of the Corsi type, whereas vi-
sual noise again had no effect. This is not a power problem as in Experiment 1, we
had a power of 0.85 in finding an effect of tapping of the same size as the one we have
found in Experiment 2. We therefore believe that in the spatial relocation task used
in Experiment 1 really no interference exists. We consider this difference between the
 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65 53

results of the two experiments as an indicator of the fact that spatial information for
objectsÕ locations is being remembered by different mechanisms than spatial informa-
tion for the locations in the spatio-temporal task of the Corsi test. Although in both
tasks location information has to be remembered, the tasks seem to tap different as-
pects of visuo-spatial short-term memory.
These results would be compatible with the VSWM model if we assumed that re-
membering of location information is a visual task that has to be distinguished from
spatial tasks, as for example the Corsi test. Then, however, a visual interference effect
should occur (Quinn & McConnell, 1996a, 1996b, 1999), i.e. visual noise should in-
terfere with the location main task. In Experiment 1 we definitely observed no such
interference. This is not compatible with the VSWM model. Nevertheless, a problem
with this result is that we argue for the acceptance of the null hypothesis. We did not
want to do so on the basis of only one study, and we therefore ran a third experiment
in which we again investigated spatial relocation performances combined with a vi-
sual secondary task.

4. Experiment 3

Although the visual secondary task in Experiment 1 had absolutely no influence

on the relocation performances, the scope of this result is still restricted, as all mem-
ory performances were high and of the same size. One might argue therefore that the
relocation task is so easy that it is not susceptible to any manipulation. We therefore
looked for a condition in which memory performances are reduced, but where a vi-
sual secondary task still does not interfere with the configuration task, independent
of the performance level.
Results from our lab suggested that the manipulation of the item content might be
suitable for this purpose. In these experiments, we tested whether memory for ob-
jectsÕ locations is a purely visual pre-semantic task. Should this be the case, the mean-
ingfulness of the presented items would not be relevant for performances. Instead,
spatial memory for artificial nonsense figures should be as good as memory for po-
sitions of real objects. However, memory for the locations of nonsense figures was
worse than memory for the locations of real objects (Zimmer, 2003). A similar result
was reported by Postma and De Haan (1996) using symbols as items. In Experiment
3, we therefore presented either abstract nonsense figures which did not depict ob-
jects or we presented pictures of real objects as in Experiment 1 (cf. Fig. 1), and
we combined the relocation task with the visual noise secondary task.
The predictions are straightforward. If the configuration of objects is recon-
structed from a perceptual memory trace stored within a visual cache, the difference
in meaningfulness should be of minor importance for performances, and visual noise
during retention should impair location memory because it overwrites the visual
cache. In contrast, if we replicate the results from Experiment 1 with this more dif-
ficult material, then the meaningfulness of the material (using nonsense figures) but
not visual noise should impair memory. Finally, a third outcome seems possible. If
in Experiment 1 the reconstruction of object-locations was supported by the use of
54 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65

semantic information, an interaction effect should occur. Object-performance should

show no interference with visual noise, whereas the performance with artificial,
meaningless figures ought to be impaired by the presence of visual noise because
their configuration can only be retrieved from the visual cache.

4.1. Method

4.1.1. Participants
Thirty-two Saarland University students took part in the experiment for payment,
again about two-third were female. None of them had participated in any of the
other experiments.

4.1.2. Material and design

Sixty-four unique artificial figures were constructed by combining abstract geo-
metrical shapes. These figures were visually clearly distinct from each other (cf.
Fig. 1, lower half), but they had no meaning. They were arbitrarily colored to make
their appearance similar to that of the objects. Sixteen configurations were con-
structed from these figures, each depicting four items, and they were assigned to
two lists of eight layouts. The figures were placed in the same layouts used for the
real objects so that two sets of configurations resulted that only differed in the de-
picted items (objects or artificial figures). As a secondary task, we used the dynamic
visual noise condition from Experiment 1. These two manipulations defined a 2 · 2
design with the factors type of item and presence of secondary task. Both factors
were manipulated within subjects to enhance power of the experiment. Subjects stud-
ied eight configurations with and eight without a secondary task (half of them with
objects and half with artificial figures). Which configurations were used in the two
conditions was counterbalanced, as was the sequence of secondary task conditions.

4.1.3. Procedure
The procedure was the same as in the visual noise condition of Experiment 1.

4.1.4. Results and discussion

We analyzed the relocalization performances as before. In Fig. 4 the proportion
of items placed in their correct positions is depicted. As one can see immediately,
the only significant effect is the type of material. A 2 · 2 analysis of variance with
the factors meaningfulness of material and presence of secondary task yielded only
a significant effect of the item type, F ð1; 31Þ ¼ 63:21, MSE ¼ 0:012, p < 0:001. The
positions of objects (0.81) were remembered better than the positions of artificial
figures (0.65). The theoretically important factor Ôpresence of secondary taskÕ had
a value of F ð1; 30Þ ¼ 0:65, p 6 0:42. Obviously there is an effect of the meaningful-
ness of the items on location memory, but not of the presence of a secondary task.
Next we had a closer look at the type of errors made. Two types of errors were
possible: (a) positions were remembered correctly but the items were confused (false
item-to-position assignment), and (b) completely wrong positions were chosen, that
were not occupied by any item (false position memory). We separately analyzed both
 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65 55

1.0

0.9

0.8
Proportion Correct

0.7

0.6

0.5

0.4

0.3
Without With Without With
Objects Nonsense Figures

Presence of Secondary Task and Type of Item

Fig. 4. Relocation performances in the objectsÕ relocation task dependent on the presence of visual noise,
and the type of items (objects, artificial figures). Note: the whiskers correspond to the 95% confidence in-
tervals of the mean.

scores. These analyses revealed, that the object advantage is based on both compo-
nents. Subjects confused objects less often than nonsense figures, given that the po-
sitions themselves were correct, F ð1; 31Þ ¼ 37:05, MSE ¼ 0:002, p < 0:001, and
furthermore they completely mislocated nonsense figures more often than objects,
F ð1; 31Þ ¼ 42:05, MSE ¼ 0:009, p < 0:001. No effect of the visual noise was observed
in any of these scores ðF < 1Þ.
We had observed the same results in a further study that was run in our labora-
tory by Wilhelm (2000). The main task of this study was comparable to the one of
Experiment 3, but as Logie (1986) had done, irrelevant pictures were presented dur-
ing retention instead of visual noise. In this experiment, we observed exactly the
same results as we obtained in Experiment 3. Location information of nonsense fig-
ures was remembered worse than that of objects, and the additional interference ma-
terial did not influence memory. We therefore conclude that neither pure location
memory nor object-to-position binding is impaired by the additional presentation
of visual material during retention, whereas in contrast, the meaningfulness of items
had a strong effect on both components.

5. Experiment 3a

Although the data of Experiment 3 were clear cut, the reader might not be com-
pletely satisfied by the experiment. One might wish to have the additional proof that
also a spatial secondary task does not impair memory for objectsÕ location with this
56 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65

new material. 3 When we planned Experiment 3, we decided to run only the visual
interference condition because according to the tripartite model the spatial location
of objects should be held within the visual and not the spatial component of the
VSWM. However after in two experiments, a visual secondary task did not interfere
with spatial memory, one might speculate that the main task may be spatial in na-
ture, and not visual. In order to test this, we decided additionally to run a spatial
tapping task with the objects and nonsense figures used in Experiment 3. If the re-
sults from Experiment 1 are replicated, spatial tapping should not impair location
memory for objects. If configurations of non-objects are remembered in the same
way, spatial memory for non-objects, too, should not be impaired by spatial tapping.
However, if the configuration of artificial figures is hold within the spatial component
of the VSWM, spatial tapping should impair memory for this purely visual material.

5.1. Method

5.1.1. Participants, design and procedure

Twenty students participated in the experiment for payment. About half of them
were male. They studied two lists of configurations constructed from the same ma-
terial as in Experiment 3. One list was studied under standard learning conditions
without tapping, the second list was studied with a spatial tapping interference dur-
ing maintenance as in Experiment 1. Each list comprised two blocks of eight config-
urations with four objects each. One block depicted objects and the other nonsense
figures. The sequence of blocks and of conditions was counterbalanced.

5.1.2. Results and discussion

In an analysis of variance for the correctly relocated figures with the factors type
of item and presence of secondary task, the type of item factor was significant. Po-
sitions of objects (0.73) were remembered better than those of arbitrary figures
(0.54), F ð1; 19Þ ¼ 142:13, MSE ¼ 0:0049, p < 0:001. The secondary task interacted
with the type of material, F ð1; 19Þ ¼ 7:12, MSE ¼ 0:010, p < 0:05. Memory for ob-
jectsÕ locations was not influenced by spatial tapping (0.73, 0.73), whereas memory
for the locations of nonsense figures was somewhat better without (0.61) than with
spatial tapping (0.48), tð19Þ ¼ 3:86, p < 0:01. The type of item effect clearly repli-
cated the results of Experiment 3, and the absence of spatial interference with spatial
memory for objects replicated Experiment 1. This again demonstrates that short-
term memory for locations of objects is not impaired by a spatial secondary task.
In contrast to objects, memory for locationsÕ of artificial figures was influenced by
spatial tapping. This might indicate that in memory for the locations of non-objects
different processes are involved than in memory for locations of objects. However,
the memory impairment by spatial tapping was much smaller ()0.13), than the influ-

3
This experiment goes back to a suggestion of one of the reviewers who emphasized that we could make
a stronger point if we would additionally exclude the possibility of a spatial interference effect. We want to
express our thanks for this hint.
 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65 57

ence of tapping on the Corsi task ()0.46). We therefore believe that the memory de-
crement was caused by distraction of general attention and not by a specific spatial
interference. We will come back to this aspect later.

6. General discussion

The aim of these experiments was to test whether temporary memory for the spa-
tial locations of objects is a ÔspatialÕ or a ÔvisualÕ task in terms of the VSWM model.
For this purpose, participants relocated objects or artificial figures, and during reten-
tion either no interference was present, or a visual secondary task was administered
(visual noise) or a spatial secondary task (spatial tapping). None of these tasks im-
paired temporary memory for objectsÕ locations. The meaningfulness of items on the
other hand was important. Memory was lower for artificial figures than for objects,
and with nonsense figures, spatial tapping had a small negative effect on memory for
locations whereas visual noise had not. In contrast, spatial tapping, but not dynamic
visual noise, caused a strong interference with the Corsi test. This was even true for
the number of correct locations independent of their temporal position in serial re-
call. This means that spatial tapping did not exclusively impair sequential informa-
tion, but spatial information was also lost if subjects tried to remember locations in a
Corsi test. Subjects did not only forget when a position was marked, they also forgot
where the position was.
The lack of an interference effect with memory for objectsÕ locations is particularly
remarkable if we consider the strong influence of tapping on the Corsi test. If inter-
ference by spatial tapping is an indicator for the involvement of a proper spatial
short-term memory, memory for the configuration of objects is qualified as visual
and not as spatial as was suggested by proponents of the VSWM model. However,
according to the model of VSWM, it follows that memory for objectsÕ locations are
provided by the visual cache. A consequence hereof should be that additional visual
input interferes with memory (Logie, 1986; Logie & Marchetti, 1991; Phillips, 1983;
Tresch et al., 1993). This was of course not the case. We therefore conclude that also
a visual cache is not used for temporary storage of spatial locations of objects. What
is the alternative?
We believe that the resistance of object-location memory against interference dur-
ing maintenance is caused by the fact that the relocation task is solved by access to
perceptual records that persist for a longer duration than the entries used in the
Corsi test. Furthermore, we assume that the visual records of objects are not regularly
served by a refresh mechanism and they are therefore not disrupted by additional in-
put during maintenance. For nonsense figures this might be partially different. Be-
cause these figures are less well remembered in long-term memory, participants
seem actively to rehearse them. As participants had occasionally told after the exper-
iment, they mentally switched from figure to figure and by this way tried to keep
alive the figureÕs appearance in consciousness. Spatial memory should profit from
this refresh mechanism, as participants need to remember the conjunction of object
and location. In contrast, the Corsi test performance always requires serial rehearsal
58 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65

of locations, and this is interrupted by a spatial secondary task. The cognitive de-
mands of these two visual working memory tasks are therefore quite different.

6.1. Short-term location memory as outcome of temporarily active episodic long-term

structures

We suggest that spatial information regarding the location of objects within con-
figurations is encoded together with the objectÕs identity during visual processing and
that this information is stored within structures that are also used for episodic long-
term memory. Location information might be stored as a configuration, i.e. a pattern
of locations. The information of the occupied location is attached to object represen-
tations which bind together the features in memory units. During test, location infor-
mation can be retrieved via access to object information; it is not necessary to get
access to this memory by temporarily tagged locations. Because the memory records
are created in the course of encoding, and as they do not need refresh, performing
secondary tasks during retention does not impair memory.
In terms of the VSWM model, one may call these records entries in an Ôepisodic
bufferÕ––a structure that was suggested by Baddeley (2000) as a further form of
working memory in addition to the phonological loop and the VSWM. All charac-
teristics that were mentioned for elements within this buffer apply to locations of ob-
jects within a configuration. Entries in the episodic buffer should be representations
that integrate information regarding an item. In the present experiment it is informa-
tion about the object, its features, and its location. Within the multi component
working memory model one can therefore explain the results by assuming that the
locations of objects are retrieved from an episodic, and not from a visual buffer,
i.e. the visual cache.
However, not only the suggestion of an episodic buffer holds that working mem-
ory consists of active long-term representations (cf. Ruchkin, Grafman, Cameron, &
Berndt, in press). The assumption of a temporarily accessible representation of visual
processing results, for example, can also be found in the work of Treisman. She sug-
gested that object files are generated during perception, and these files collect object
features including position. The object files constitute objects tokens in memory (e.g.
Treisman & Kanwisher, 1998). Schneider (1999) explicitly discussed these object files
as elements of working memory. Similarly, Cowan (1999) assumed that working
memory is the active part of long-term memory. Finally, in the context of cognitive
neuroscience, Murre (1997) suggested that features of an episode represented in a
distributed fashion are temporarily bound by the hippocampus until the Ôconjunctive
ensemblesÕ are consolidated (cf. also Paller, 2000). These ensembles with an interme-
diate persistence are the basis for the enduring long-term memory and they ÔsurviveÕ
without refreshment for a medium length of time. Therefore, they do not need active
rehearsal during maintenance. All these assumptions would predict results like those
we have reported and they are in good agreement with our suggestion that the short-
term relocation task is solved by using more permanent memory traces.
However, there are even more arguments in favor of this idea. For example, long-
term memory for objectsÕ locations is quite good (e.g., Mandler, Seegmiller, & Day,
 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65 59

1977; Pezdek et al., 1986; Zimmer & de Vega, 1996), and there is no reason for not
using these Ôlong-term structuresÕ with short retention intervals. Furthermore, the
quality of long-term memory is usually a function of the meaningfulness of items,
and we have observed the same in a spatial long-term memory task (Zimmer,
2003). Our short-term relocation task was strongly influenced by the meaningfulness
of the pictures as well (Experiments 3 and 3a). The task therefore does not seem to be
a low level, purely visual, pre-semantic one.
Additionally, if a transient short-term memory is used for relocating the items,
one should expect a clear influence of the length of the retention interval on location
memory. In fact, the effect of a delay seems to be rather small. Kessels, Postma, Wes-
ter, and de Haan (2000) observed lower memory performances with a long delay (180
s) than with a zero delay, but compared to the large difference between these two in-
tervals, the effect was small. Morris (1987) studied memory for the positions of cir-
cles and he varied the retention interval from 0 to 20 s. He did not observe any effect.
We found the same result in a series of experiments using an S1–S2 paradigm (Zim-
mer & Lehnert, submitted for publication). As expected, we observed robust effects
of spatially incongruent configurations on object memory although this information
was irrelevant. However, the negative effect of a spatially incongruent target stimulus
was neither influenced by additional visual input that was presented during the reten-
tion interval, nor was it diminished if we extended the retention interval from 1 to
10 s. This is the same pattern than the one observed with the relocation task.
Finally, in brain imaging and evoked potential studies correlates of neural activ-
ities were found, which suggest that partly the same neural structures are active in
spatial long-term tasks (e.g., Moscovitch, Kapur, K€ ohler, & Houle, 1995; Rolke,
Heil, Hennighausen, H€ aussler, & R€ osler, 2000; R€
osler, Heil, & Hennighausen,
1995) and in so-called short-term memory tasks (Mecklinger, 1998, 1999; Mecklinger
& M€ uller, 1996; Mecklinger & Pfeifer, 1996). Long-term and short-term representa-
tions for configurations of objects may therefore not differ in the supporting neural
substrate, but only in the status of activation and in the processes that work on these
structures (Rolke et al., 2000). The contrast therefore does not exist between short
and long-term memory but between object and spatial information. It is very likely
that in long- as well as in short-term memory, object and spatial information are sep-
arate components which can be independently addressed but which are bound within
episodic memory entries (e.g., Mecklinger & Meinshausen, 1998; Moscovitch et al.,
1995).
The fact that nonsense figures behaved differently is not a contradiction to this po-
sition. Long-term memory can be activated or deactivated. The status of activation is
not important for objects because it is easy to reactivate them. However, nonsense
figures are less well encoded and they can easily be confused. This ÔnoiseÕ within
the memory representation makes it difficult to reintegrate a nonsense figure during
retrieval. If it is difficult to retrieve an item, keeping figures active in memory is an
advantage. If such a refresh process is hindered, memory should be impaired. Be-
cause spatial tapping demands attention, it can reduce the effectiveness of the refresh
mechanism to some extent, and by this it also impairs memory for the locations of
the individual nonsense figures. According with this, after the experiment, subjects
60 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65

reported that they were more active while remembering nonsense figures than while
processing objects. Additionally, note that the interference effect of spatial tapping
was clearly higher for memory of locations in Corsi sequences than in configurations
of nonsense figures. We therefore believe that the interference of spatial tapping on
memory for locations of nonsense figures is not a specific spatial effect but a more
general effect.

6.2. Short-term spatio-temporal memory as outcome of spatial marking served by

spatial attention

Given this position, the question arises why the Corsi test was so strongly im-
paired by the spatial secondary task but not by the visual one. This pattern of inter-
ference effects is compatible with the assumption that spatio-temporal and visual
components within the VSWM differ, and that the Corsi- and the tapping task par-
tially share the same limited component or the same information. In fact, in both
tasks changes of locations over time have to be processed. Additionally, in the spatial
task only weak item cues exist to support memory. We need to consider both aspects.
In the Corsi test, participants are required to remember a sequence of spatial lo-
cations in a homogeneous field of items, e.g. circles, which are only temporarily
marked (cf. also Berch, Krikorian, & Huha, 1998). The target locations can only
be accessed by a conjunction of spatial coordinates and a temporal index marking
the location. A closer look at the typical so-called spatial tasks reveals that these
features are present in most, if not in all, tasks called spatial in the narrower sense
within the VSWM model. The standard spatial tasks include a temporal and/or a
sequential component within a spatial medium, e.g., the Brooks matrix task, the
Corsi test, spatial tapping, or the spatial span task. Furthermore, contrary to the ob-
ject-location task, in the Corsi task subjects cannot obtain access to any location if
not by its spatial coordinates. None of the locations have any persisting discrimina-
tive non-spatial features. A memory entry is unique only due to its temporal mark-
ing. Therefore, a conjunction of a spatial location and a temporal index must be
held. Probably this conjunction cannot be instantly stored, and for this reason, re-
hearsal is necessary to avoid the fading of information.
Rehearsal can be thought of as a shift of spatial attention over target positions
(Awh & Jonides, 2001). This necessity of spatial rehearsal for maintenance of infor-
mation is the reason for memory impairment on the Corsi test by a secondary task
during the retention interval. During spatial tapping spatial attention is directed to
locations other than the target positions, and it is therefore not possible to shift atten-
tion over target locations during maintenance in order to rehearse the spatial loca-
tions. This dividing of attention over locations causes interference. It is involved in
the control of spatial movement, in locating individual stimuli which are spatially
distributed (e.g., Smyth & Scholey, 1994), and also in perceiving or rather visual
tracking of an isolated moving stimulus (e.g. Quinn & McConnell, 1996b). Jones,
Farrand, Stuart, and Morris (1995) took the position that shifting (spatial) attention
is a supramodal process shared by many modalities. The involvement of spatial atten-
tion may also explain the apparently deviating result of Tresch et al. (1993). In their
 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65 61

experiment, location memory (remembering an isolated small dot on an otherwise

empty screen) was impaired by a movement detection task (detecting a stationary star
among moving stars and indicating its position). For both tasks, spatial attention is
probably necessary. It is likely that the exact position of an isolated dot within a wide
space can only be retained if attention is focused on the target position, and in order
to indicate a target position, spatial attention has to be focused, too. In general we
therefore believe that a spatial working memory task and a secondary task during
maintenance interfere with each other if the main task makes it necessary to orient
spatial attention in order to keep location memory ÔaliveÕ and the secondary task ori-
ents attention to different positions, be it voluntarily or involuntarily.

6.3. Final remarks

Before closing, we want to add some final remarks regarding the two tasks we
have used, and regarding research on visual working memory in general.
First, we have to say that the spatial relocation task and the Corsi test, of course
differ in many ways. It is not only the fact that in one condition spatial locations of
objects are to be remembered, whereas in the other it is a spatio-temporal sequence.
There are other differences as well. For example, in the location task, objects were
used instead of dots, and the items were presented simultaneously instead of in a se-
quential manner. Therefore, we cannot be sure whether the variables we have dis-
cussed as causing the different results are truly critical. However, we think that
they are the most plausible variables. Apart from the discussed differences, we only
see one further variable that may be considered as relevant. It is the simultaneous
presentation of items in the object location task versus the sequential presentation
in the Corsi task. This variable has of course an influence on spatial processing,
but we believe simultaneity per se did not cause the differences in the two tasks.
The important stimulus factor in the object location task is the possibility men-
tally to represent a spatial configuration of interrelated objects. For this purpose,
more than one object must be present within the same ÔsceneÕ. This is clearly the case
if several objects are simultaneously presented, but with explicit instructions as we
had used, such an integrated representation is also generated from a sequential pre-
sentation of objects. For example, Pazzaglia and Cornoldi (1999) contrasted different
visual and spatial tasks with simultaneous and sequential characteristics. They ob-
served a clear difference only between verbal and visual Brooks tasks on the one
hand and a verbal or serial secondary task on the other hand. Except for this differ-
ence, simultaneous and sequential tasks behaved in a similar way. Schumann-
Hengsteler compared spatial reconstruction under simultaneous, sequential or
cumulative presentation conditions––all data are cited from Schumann-Hengsteler
(1995). She did not find any influence of simultaneity on position memory with either
adult participants (Schumann-Hengsteler et al., 1993), or 5–10-year-old children
(Schumann-Hengsteler, 1992). Only if identical objects were presented did a simulta-
neous presentation have an advantage over a sequential presentation (Schumann-
Hengsteler et al., 1992). Finally, in our lab we investigated spatial learning from
maps that were presented piecemeal as fragments compared to the presentation of
62 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65

the complete map. Again, both presentation conditions yielded comparable results
(Zimmer, submitted for publication). We therefore conclude that it is only under spe-
cific conditions that simultaneous presentations may have an advantage over sequen-
tial presentations. It therefore seems quite unlikely that the way of presentation
caused the difference between the spatial location and the Corsi task.
Additionally, it should be noted that the two procedures were carried out as they
typically are, and in a way they have been used in many experiments to investigate
visuo-spatial short-term memory. Our aim was to test whether the objectsÕ location
task is visual or spatial in terms of the VSWM model. The obtained results show that
short-term spatial location memory is probably not served by the spatial component
which is used in the Corsi task, but more likely by temporarily active episodic mem-
ory traces. Only if one regards these records as the visual cache and one assumes that
this ÔbufferÕ is not overwritten by additional input, the results are compatible with the
VSWM model.
In closing, we want to draw the readersÕ attention to the variability of paradigms
that were used to investigate visuo-spatial memory. When we planned our studies,
we recognized that both tasks had been realized in many different ways, and these
differences were not considered relevant. Sometimes VSWM was investigated by ob-
ject location tasks, and sometimes by a Corsi test, and both tasks were established
with a huge variation of parameters. To give a few examples, Luck and Vogel
(1997) simultaneously presented between 3 and 8 simple figures in an S1–S2 para-
digm. S1 was visible for 200 ms and after a retention interval of 900 ms, S2 was pre-
sented. Awh, Jonides, and Reuter-Lorenz (1998) presented S1 for a similar time of
400 ms, however, they presented S2 after 5000 ms. Logie and Marchetti (1991) used
a sequential presentation of color patches at a rate of one item per second and a re-
tention interval of 10 s. Mecklinger and M€ uller (1996) also presented objects sequen-
tially, but they showed 8 items for 2.5 s per object with a 30 s retention interval.
Postma and De Haan (1996) used a 30 s retention interval, but they presented
4–10 objects (symbols) simultaneously. Mecklinger and Meinshausen (1998) sequen-
tially presented 4 objects (2 s/object) with a 3 s retention interval, whereas Mecklin-
ger and Pfeifer (1996) simultaneously showed up to five objects for 1800 ms with a
retention interval of 5000 ms. These procedural differences were not discussed in any
of these experiments. A similar variability can be found with the Corsi test as the
overviews in Berch et al. (1998) and Fischer (2000) demonstrate.
Given that in the models of VSWM nothing is said about the theoretical impor-
tance of these variables, this lack of differentiation is not carelessness from the exper-
imentersÕ side. However, different short-term memory tasks may have very different
task requirements, as our results have demonstrated, and we should therefore be more
cautious with respect to the task and the timing we choose to investigate VSWM.

Acknowledgements

This research was supported by a grant from the Deutsche Forschungsgemeins-

chaft in a Special Collaborative Research Group on Resource Adaptive Cognitive
 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65 63

Processes (SFB 378). We thank Ullrich Ecker, Ellen Koch, Silke Neiss and Nadine
Schubel for their assistance in performing the experiments, and Carsten Wilhelm for
preparing the artificial figures. We acknowledge the helpful suggestions of Albert
Postma and of the anonymous reviewers.

References

Awh, E., & Jonides, J. (2001). Overlapping mechanisms of attention and spatial working memory. Trends
in Cognitive Sciences, 5, 119–126.
Awh, E., Jonides, J., & Reuter-Lorenz, P. A. (1998). Rehearsal in spatial working memory. Journal of
Experimental Psychology: Human Perception and Performance, 24, 780–790.
Baddeley, A. D. (2000). The episodic buffer, a new component of working memory. Trends in Cognitive
Sciences, 4, 417–423.
Baddeley, A. D., & Hitch, G. (1974). Working memory. In G. A. Bower (Ed.), Recent advances in learning
and motivation (Vol. 8, pp. 47–90). New York: Academic Press.
Baddeley, A. D., & Lieberman, K. (1980). Spatial working memory. In R. Nickerson (Ed.), Attention and
performance VIII (pp. 521–539). Hillsdale: Lawrence Erlbaum.
Berch, D. B., Krikorian, R., & Huha, E. M. (1998). The Corsi block-tapping task: Methodological and
theoretical considerations. Brain and Cognition, 38, 317–338.
Chalfonte, B. L., & Johnson, M. K. (1996). Feature memory and binding in young and older adults.
Memory and Cognition, 24, 403–416.
Cowan, N. (1999). An embedded-processes model of working memory. In A. Miyake, & P. Shah (Eds.),
Models of working memory (pp. 62–101). Cambridge: University Press.
Fischer, M. H. (2000). Probing spatial working memory with the Corsi Blocks task. Brain and Cognition,
45, 143–154.
Gathercole, S. E., & Baddeley, A. D. (1993). Working memory and language. Hove: Lawrence Erlbaum.
Goodale, M. A., & Humphrey, G. K. (1998). The objects of action and perception. Cognition, 67, 181–207.
Hecker, R., & Mapperson, B. (1997). Dissociation of visual and spatial processing in working memory.
Neuropsychologia, 35, 599–603.
Jiang, Y., Olson, I. R., & Chun, M. M. (2000). Organization of visual short-term memory. Journal of
Experimental Psychology: Learning, Memory, and Cognition, 26, 683–702.
Jones, D., Farrand, P., Stuart, G., & Morris, N. (1995). Functional equivalence of verbal and spatial
information in serial short-term memory. Journal of Experimental Psychology: Learning, Memory, and
Cognition, 21, 1008–1018.
Kemps, E. (1999). Effects of complexity on visuo-spatial working memory. European Journal of Cognitive
Psychology, 11, 335–356.
Kessels, R. P., Postma, A., & de Haan, E. H. F. (1999). P and M channel-specific interference in the what
and where pathway. Neuroreport, 10, 3765–3767.
Kessels, R. P., Postma, A., Wester, A. J., & de Haan, E. H. F. (2000). Memory for object locations in
KorsakoffÕs amnesia. Cortex, 36, 47–57.
Logie, R. H. (1986). Visuo-spatial processing in working memory. Quarterly Journal of Experimental
Psychology, 38A, 229–247.
Logie, R. H. (1995). Visuo-spatial working memory. Hove: Lawrence Erlbaum.
Logie, R. H., & Marchetti, C. (1991). Visuo-spatial working memory: Visual, spatial or central executive?
In R. H. Logie, & M. Denis (Eds.), Mental images in human cognition. Amsterdam: North-Holland.
Logie, R. H., & Pearson, D. (1997). The inner eye and the inner scribe of visuo-spatial working memory:
Evidence from developmental fractionation. European Journal of Cognitive Psychology, 9, 241–257.
Logie, R. H., Zucco, G. M., & Baddeley, A. D. (1990). Interference with visual short-term memory. Acta
Psychologica, 75, 55–74.
Luck, S., & Vogel, E. (1997). The capacity of visual working memory for features and conjunctions.
Nature, 390, 279–281.
64 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65

Mandler, J., Seegmiller, D., & Day, J. (1977). On the coding of spatial information. Memory & Cognition,
5, 10–16.
Mecklinger, A. (1998). On the modularity of recognition memory for object form and spatial location: A
topographic ERP analysis. Neuropsychologia, 36(5), 441–460.
Mecklinger, A. (1999). Das Erinnern von Orten und Objekten. G€ ottingen: Hogrefe.
Mecklinger, A., & Meinshausen, R. M. (1998). Recognition memory for object form and object location:
An event-related potential study. Memory and Cognition, 26, 1068–1088.
Mecklinger, A., & M€ uller, N. (1996). Dissociations in the processing of ‘‘what’’ and ‘‘where’’ information
in working memory: An event-related potential analysis. Journal of Cognitive Neuroscience, 8, 453–473.
Mecklinger, A., & Pfeifer, E. (1996). Event-related potentials reveal topographical and temporal distinct
neuronal activation patterns for spatial and object working memory. Cognitive Brain Research, 4, 211–
224.
Morris, N. (1987). Exploring the visuo-spatial sketch pad. Quarterly Journal of Experimental Psychology,
39A, 409–430.
Moscovitch, M., Kapur, S., K€ ohler, S., & Houle, S. (1995). Distinct neural correlates of visual long-term
memory for spatial location and object identity: A positron emission tomography study in humans.
Proceedings of the National Academy of Sciences USA, 92, 3721–3725.
Murre, J. M. (1997). Implicit and explicit memory in amnesia: Some explanations and predictions by the
Trace Link Model. Memory, 5, 213–232.
Naveh-Benjamin, M. (1987). Coding of spatial location information: An automatic process? Journal of
Experimental Psychology: Learning, Memory, and Cognition, 13, 595–605.
Paller, K. A. (2000). Neurocognitive foundations of human memory. In D. L. Medin (Ed.), The psychology
of learning and motivation (pp. 121–145). San Diego: Academic Press.
Pazzaglia, F., & Cornoldi, C. (1999). The role of distinct components of visuo-spatial working memory in
the processing of texts. Memory, 7, 19–41.
Pezdek, K., Roman, Z., & Sobolik, K. G. (1986). Spatial memory for objects and words. Journal of
Experimental Psychology, 12, 530–537.
Phillips, W. A. (1983). Short-term visual memory. Philosophical Transactions of the Royal Society of
London B, 302, 295–309.
Postma, A., & De Haan, E. H. F. (1996). What was where? Memory for object locations. Quarterly
Journal of Experimental Psychology: Human Experimental Psychology, 49A, 178–199.
Quinn, J. G., & McConnell, J. (1996a). Irrelevant pictures in visual working memory. Quarterly Journal of
Experimental Psychology: Human Experimental Psychology, 49A, 200–215.
Quinn, J. G., & McConnell, J. (1996b). Interference at the encoding and maintenance of visual
information. Psychologische Beitr€age, 38, 343–354.
Quinn, J. G., & McConnell, J. (1999). Manipulation of interference in the passive visual store. European
Journal of Cognitive Psychology, 11, 373–389.
Quinn, J. G., & Ralston, G. E. (1986). Movement and attention in visual working memory. Quarterly
Journal of Experimental Psychology, 38A, 689–703.
R€
osler, F., Heil, M., & Hennighausen, E. (1995). Distinct cortical activation patterns during long-term
memory retrieval of verbal, spatial, and color information. Journal of Cognitive Neuroscience, 7,
51–65.
Rolke, B., Heil, M., Hennighausen, W., H€aussler, C., & R€ osler, F. (2000). Topography of brain electrical
activity dissociates the sequential order transformation of verbal versus spatial information in humans.
Neuroscience Letters, 282, 81–84.
Ruchkin, D. S., Grafman, J., Cameron, K., & Berndt, R. S. (in press). Working memory retention systems:
A state of activated long-term memory. Behavioral and Brain Sciences.
Santa, J. L. (1977). Spatial transformations of words and pictures. Journal of Experimental Psychology:
Human Learning and Memory, 3, 418–427.
Schneider, W. X. (1999). Visual-spatial working memory, attention, and scene representation: A neuro-
cognitive theory. Psychological Research, 62, 220–236.
Schumann-Hengsteler, R. (1995). Die Entwicklung des visuell-r€aumlichen Ged€achtnisses. G€ ottingen:
Hogrefe.
 H.D. Zimmer et al. / Acta Psychologica 114 (2003) 41–65 65

Smith, E. E., & Jonides, J. (1997). Working memory: A view from neuroimaging. Cognitive Psychology,
33, 5–42.
Smith, E. E., Jonides, J., Koeppe, R. A., Awh, E., Schumacher, E. H., & Minoshima, S. (1995). Spatial
versus object working memory: PET investigations. Journal of Cognitive Neuroscience, 7, 337–356.
Smyth, M. M., Pearson, N. A., & Pendleton, L. R. (1988). Movement and working memory: Patterns and
positions in space. Quarterly Journal of Experimental Psychology, 40A, 497–514.
Smyth, M. M., & Scholey, K. A. (1994). Interference in immediate spatial memory. Memory & Cognition,
22, 1–13.
Toms, M., Morris, N., & Foley, P. (1994). Characteristics of visual interference with visuospatial working
memory. British Journal of Psychology, 85, 131–144.
Treisman, A. M., & Kanwisher, N. G. (1998). Perceiving visually presented objects: recognition, awareness
and modularity. Cognitive Neuroscience, 8, 218–226.
Tresch, M. C., Sinnamon, H. M., & Seamon, J. G. (1993). Double dissociation of spatial and object visual
memory: Evidence from selective interference in intact human subjects. Neuropsychologia, 31, 211–219.
Wilhelm, C. (2000). Der Irrelevant Picture Effekt beim r€aumlichen Erinnern. Unpublished diploma thesis.
Saarbr€ucken: Department of Psychology, Saarland University.
Zimmer, H. D. (1998). Spatial information with pictures and words in visual short-term memory.
Psychological Research, 61, 277–284.
Zimmer, H. D. (2003). Spatial long-term memory and encoding of inter-item spatial relations, unpublished
data.
Zimmer, H. D. (submitted for publication). The construction of mental maps based on a fragmentary view
of physical maps.
Zimmer, H. D., & de Vega, M. (1996). The medium and the message in spatial communication. In T.
Ensink, & C. Sauer (Eds.), Researching technical documents (pp. 153–180). Groningen: University of
Groningen.
Zimmer H. D., & Lehnert, G. (submitted for publication). Spatial mismatch effects in the S1–S2 paradigm
are caused by global configuration memory and not by the visual cache.