Cognitive Brain Research 18 (2003) 102 – 105 www.elsevier.com/locate/cogbrainres
Department of Psychology, University of British Columbia, 2136 West Mall, Vancouver, BC, Canada V6T 1Z4
Accepted 27 August 2003
Abstract
Recent evidence indicates that reflexive shifts in spatial attention with eye movements (overt orienting) and without eye movements
(covert orienting) can be dissociated [J. Exp. Psychol. Hum. Percept. Perform., in press]. Here, we show that a similar dissociation exists for voluntary shifts in overt and covert attention. Our study is consistent with general theories of attention that assume bottom-up (reflexive) processes and top-down (voluntary) processes converge on a common neural architecture.
D
2003 Elsevier B.V. All rights reserved.
Theme: Neural basis of behavior
Topic: Cognition
Keywords: Covert visual attention; Eye movements; Voluntary and reflexive attention
Shifts in attention need not entail an overt shift of the eyes
[15] . Nevertheless, spatial attention and the eyes often
move about the environment in tandem. For instance, an abrupt onset in the visual periphery can reflexively ‘‘capture’’ both attention
[20]
and the eyes
[19] . What is not
known is whether both the eyes and attention are attracted to salient external events because the two forms of orienting are linked by a common neural architecture, or because they are independent with each being activated by an abrupt onset. While many investigators have suggested that the former is the case (e.g.,
Refs. [2,4,11,14,17] ), we have
recently demonstrated a double dissociation between eye movements and covert attention with respect to the reflexive effects of abrupt onsets, suggesting that the two forms of orienting are independent
[7] . In the present study, we asked
whether a similar separation occurs when overt and covert orienting are activated volitionally, that is, in response to the internal goals and expectations of the individual. There are a number of reasons to think that the effects of volitional attention may differ from reflexive attention. For instance, in contrast to reflexive attention, volitional orienting occurs more slowly
[12] , is more vulnerable to inhibition
[1] , may
* Corresponding author. Tel.: +1-604-822-0069; fax: +1-604-822-
6923.
E-mail address: AmeliaHunt@imag.net (A.R. Hunt).
0926-6410/$ - see front matter
D
2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.cogbrainres.2003.08.006
involve important differences in the representation of space
[16]
and is extremely sensitive to dual task demands
[8] .
This last point is particularly relevant because dual task paradigms were used in the past to test the link between overt and covert volitional orienting directly
[9,10] . The
logic of these investigations was straightforward: If the two forms of orienting are linked, then activation of one should activate the other. That is, if preparing an eye movement to a location causes attention to be allocated there, then stimulus detection at that location should be facilitated. Conversely, if allocating attention to a location causes an eye movement to be prepared there, then saccadic reaction time (RT) to that location should be facilitated.
No such link between volitional attention and eye movements was observed. When the primary task was to prepare an eye movement to a particular location, people were fastest to move their eyes there (indicating that an eye movement had been prepared in accordance with the primary task); but the secondary task of detecting a visual probe with a key press was unaffected by where the eyes were prepared to move. Conversely, when the primary task was to attend covertly to a particular location, people were fastest to make a keypress to a visual target detected there
(indicating that covert attention had been allocated in accordance with the primary task); but the secondary task of moving the eyes was unaffected by where covert attention was allocated. While these results suggest that voli-

A.R. Hunt, A. Kingstone / Cognitive Brain Research 18 (2003) 102–105 tional eye movements and attention are independent, this conclusion is premature because speeded responses were required for both the primary and secondary responses. This is potentially problematic because when two different response tasks are speeded, subjects may sacrifice speed in the secondary task to maximize performance in the primary task, thereby masking any effects in the slower secondary task. The results
[9,10]
agree with this interpretation: While the primary task always showed the expected effects, the secondary task did not.
The present study addressed this issue by requiring that only one task be speeded. This change should eliminate any performance dependencies between the two tasks, because attention can be divided between two tasks when one response is speeded and the other is not
[13] . We ran
two versions of the experiment: In one, the primary task was speeded and the secondary task was not; in the other, the secondary task was speeded and the primary task was not. This procedure removes any confound between the
103 primary task and the task that was speeded, allowing us to assess directly whether speed alone produces a performance tradeoff.
We tested two groups of 25 undergraduate subjects. For one group, the primary task was to prepare an eye movement to one of two peripheral locations (Eye Movement
(EM) Group). On most trials (70%), the eye movement response would be signaled, although, occasionally, an eye movement to the other location would be signaled (10%) or a visual probe would appear briefly at one of the two locations (20%). When the visual probe appeared, subjects indicated with an unspeeded keypress whether the probe was a vertical or horizontal line.
For the second group, these two tasks were reversed.
Now, the primary task was to attend covertly to one of two peripheral locations and make a unspeeded keypress indicating whether the visual target was vertical or horizontal
(Keypress (KP) Group). On most trials, a target would appear at the attended location (70%), although, occasion-
Fig. 1. (A) Sequence of trial events. Trials began with the onset of a central fixation flanked 4 j to the left and right by two 2.4
j circles. The auditory directional precue was a male voice saying ‘‘left’’ or ‘‘right’’, followed by either an auditory (saccade task) or a visual (manual task) target. On auditory target trials a female voice said ‘‘left’’ or ‘‘right’’, instructing the subject to execute a saccade in that direction as quickly as possible. On visual trials, a line appeared inside one of the two circles and was then masked with random lines. The subject remained fixated on the center and pressed one button if the masked target was horizontal and the other if it was vertical. The target-mask SOA was set to maintain individual subject performance between 70% and 80%. For the EM group, the auditory target appeared on 80% of the trials, and 87.5% of the time it was valid. For the 20% of trials on which the visual target appeared, it was in the cued location on half the trials. This pattern of contingencies was reversed for the KP group: the visual target appeared on 80% of the trials, and 87.5% of the time it was valid; for the 20% of the trials where the auditory target occurred, the cue was not predictive. Subjects were well-practiced at performing both tasks alone
(120 trials each) and in combination (240 trials). Before each of the test blocks (4 blocks of 120 trials), subjects were reminded of the frequency of each trial type and the cue-target contingencies. (B) Results from the two groups. (1) Reaction time (saccade task) and error rate (manual task) in the EM group, for whom the saccade task was primary and precues were predictive of saccade location but not manual target location. (2) Reaction time (saccade task) and error rate (manual task) in the KP group, for whom the manual task was primary and precues were predictive of manual target location but not saccade location. One subject failed to perform the task and was excluded.

104 A.R. Hunt, A. Kingstone / Cognitive Brain Research 18 (2003) 102–105 ally, it would appear at the other location (10%), or a speeded eye movement to one of the two locations would be signaled (10%). Details are presented in
Fig. 1A .
Fig. 1B
shows mean eye movement RTs as a function of cue condition (valid or invalid). One-tailed paired t tests were used as the predictions of overt/covert coactivation were clear, and it was important to maximize power to detect possible cueing effects in both the primary and the secondary tasks. The goal of the primary saccade task was to motivate subjects in the EM group to prepare a saccade in the direction commanded by the cue. The results indicate that subjects were indeed using the cue in this manner: RT was 343 ms on valid trials and 477 ms on invalid trials, a difference that was significant, t (24) = 6.84, p < 0.001. Eye movement errors were consistent with the RTs, with fewer errors on valid than invalid trials (16% versus 25%), t (24) = 3.75, p < 0.001, again demonstrating that an eye movement had been prepared to the cued location. Did this eye movement preparation elicit a covert shift of attention to the prepared location? It did not. Visual discrimination of a target probe performance was, if anything, worse at the location toward which an eye movement had been prepared
(71% correct at the valid location) than the opposing location (82% correct at the invalid location), although it was not significant, t (24) = 1.83. Covert orienting should improve visual discrimination, not worsen it
[3] .
For the KP group, the visual discrimination task was primary, and, now, visual discrimination was enhanced at the location cued by the male voice (79% at the valid location versus 73% at the invalid location, t (23) = 1.99, p < 0.05). Did this covert orienting result in an eye movement being prepared to the attended (cued) location? It did not. Eye movements were equivalent at the attended (483 ms) and unattended (493 ms) locations, t (23) = 1.52. Eye movement errors also did not vary as a function of attended versus unattended locations (16% versus 18%, respectively, t (23) = 0.77).
1
Together, the data suggest that volitional attention and eye movements are independent. That is, the EM group showed that preparing to move one’s eyes to a particular location does not result in covert attention being allocated to that location. Likewise, the KP group showed that allocation of covert attention to a particular location does not result in the eyes being prepared to move to that location. This dovetails with the only other direct tests of the relation between volitional attention and eye movements
[8,9] , see
also
Ref. [18] . We had questioned these previous failures to
show links between volitional attention and the eye movement system because of a possible tradeoff between speeded tasks. But even when this potential problem is corrected, and even when one-tailed t tests are performed to capture any coactivation, our data converges on the same conclusion—volitional attention and eye movements are functionally independent.
Recently, we found that reflexive attention and eye movement processes are independent
[7] . When the findings
of that investigation are combined with the present study, one is led to the conclusion that attention and eye movements are independent whether they are activated reflexively or volitionally. This conclusion is particularly timely because it provides a strong foundation for recent theories of attention that are grounded on the assumption that bottom-up (reflexive) or top-down (volitional) processes have access to the same cognitive and neural architectures that subserve attention and eye movements (e.g.
Refs. [5,6] ).
In other words, bottom-up and top-down control may differ greatly, but they access and control the same independent covert and overt attentional systems.
Acknowledgements
This research was funded through grants to A.R. Hunt and to A. Kingstone from the Natural Sciences and
Engineering Research Council of Canada and grants to A.
Kingstone from the Michael Smith Foundation for Health
Research and the Human Frontiers Science Program. Data from this study were reported at the 8th Annual Meeting of the Canadian Society for Brain, Behaviour, and Cognitive
Science, in Ottawa, Ontario (Ennis and Kingstone, 1998).
We are extremely grateful to Tricia Ennis for data collection and initial work on the manuscript.
1
One might speculate that cueing effects were masked in the secondary task because it involves an element of unexpectedness that the primary task does not. If this were the case, one would predict reduced cueing effects when the task on the current trial is different from the immediately preceding trial relative to when it is repeated. We examined the effect of switching and repeating tasks on the magnitude of cueing effects for primary and secondary tasks. The data from two subjects in the EM group and one from the KP group had to be removed from analysis due to missing data in some of the cells. A 2 2 ANOVA (cued/uncued repeat/ switch) on each task did not confirm the prediction: Cueing effects did not increase when the task switched. There was a significant interaction for the saccade task when it was primary, but the direction was opposite to what was predicted, with the cueing effect being larger following a switch trial than a repeat trial, F (1,22) = 5.97, p < 0.05. For the secondary task and the primary manual task, the interaction was not significant (all p ’s>0.30).
References
[1] M.L. Cheal, D.R. Lyon, Central and peripheral precuing of forcedchoice discrimination, Q. J. Exp. Psychol., A 43 (1991) 859 – 880.
[2] M.H. Grosbach, T. Paus, Transcranial magnetic stimulation of the human frontal eye field: effects on visual perception and attention,
J. Cogn. Neurosci. 14 (2002) 1109 – 1120.
[3] H.L. Hawkins, S.A. Hillyard, S.J. Luck, M. Mouloua, C.J. Woodword, D.P. Woodword, Visual attention modulates signal detectability, J. Exp. Psychol. Hum. Percept. Perform. 16 (1990) 802 – 811.
[4] J.E. Hoffman, B. Subramanian, The role of visual attention in saccadic eye movements, Percept. Psychophys. 57 (1995) 787 – 795.
[5] J.B. Hopfinger, G.R. Mangun, Reflexive attention modulates processing of visual stimuli in human extrastriate cortex, Psychol. Sci. 9
(1998) 441 – 447.
[6] J.B. Hopfinger, M.H. Buonocore, G.R. Mangun, The neural mecha-

A.R. Hunt, A. Kingstone / Cognitive Brain Research 18 (2003) 102–105 nisms of top-down attentional control, Nat. Neurosci. 3 (2000 Mar.)
284 – 291.
[7] A.R. Hunt, A. Kingstone, Inhibition of return: dissociating attentional and oculomotor components. J. Exp. Psychol. Hum. Percept. Perform.
2003.
[8] J. Jonides, Voluntary versus automatic control over the mind’s eye’s movement, in: J.B. Long, A.D. Baddeley (Eds.), Attention and Performance, vol. IX, Erlbaum, Hillsdale, NJ, 1981, pp. 187 – 203.
[9] R.M. Klein, Does oculomotor readiness mediate cognitive control of visual attention?, in: R. Nickerson (Ed.), Attention and Performance, vol. VIII, Erlbaum, Hillsdale, NJ, 1980, pp. 259 – 276.
[10] R.M. Klein, A. Pontefract, Does oculomotor readiness mediate cognitive control of visual attention? Revisited!, in: C. Umilta, M. Moscovitch (Eds.), Attention and Performance, vol. XV, MIT Press,
Cambridge, MA, 1994, pp. 333 – 350.
[11] T. Moore, M. Fallah, Control of eye movements and spatial attention,
Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 1273 – 1276.
[12] H.J. Mu¨ller, P.M. Rabbitt, Reflexive and voluntary orienting of visual attention: time course of activation and resistance to interruption.
J. Exp. Psychol. Hum. Percept. Perform. 15, 315 – 330.
[13] H. Pashler, Dissociations and dependencies between speed and accu-
105 racy: evidence for a two-component theory of divided attention in simple tasks, Cogn. Psychol. 21 (1989) 469 – 514.
[14] R.J. Perry, S. Zeki, The neurology of saccades and covert shifts of attention: an event-related fMRI study, Brain 123 (2000) 2273 – 2288.
[15] M.I. Posner, Orienting of attention, Q. J. Exp. Psychol. 32 (1980)
3 – 25.
[16] P.A. Reuter-Lorenz, R. Fendrich, Oculomotor readiness and covert orienting: differences between central and peripheral cues, Percept.
Psychophys. 52 (1992) 336 – 344.
[17] G. Rizzolatti, L. Riggio, I. Dascola, C. Umilta, Reorienting attention across horizontal and vertical meridians: evidence in favor of a premotor theory of attention, Neuropsychologia 25 (1987) 31 – 40.
[18] P. Sumner, T. Adamjee, J.D. Mollon, Signals invisible to the collicular and magnocellular pathways can capture attention, Curr. Biol. 12
(2002) 1312 – 1316.
[19] J. Theeuwes, A.F. Kramer, S. Hahn, D.E. Irwin, Our eyes do not always go where we want them to go: capture of the eyes by new objects, Psychol. Sci. 9 (1998) 379 – 385.
[20] S. Yantis, J. Jonides, Abrupt visual onsets and selective attention: evidence from visual search, J. Exp. Psychol. Hum. Percept. Perform.
10 (1984) 601 – 621.