UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Rethinking human attention and its components Ristic, Jelena 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2007-267820.pdf [ 8.74MB ]
Metadata
JSON: 831-1.0100574.json
JSON-LD: 831-1.0100574-ld.json
RDF/XML (Pretty): 831-1.0100574-rdf.xml
RDF/JSON: 831-1.0100574-rdf.json
Turtle: 831-1.0100574-turtle.txt
N-Triples: 831-1.0100574-rdf-ntriples.txt
Original Record: 831-1.0100574-source.json
Full Text
831-1.0100574-fulltext.txt
Citation
831-1.0100574.ris

Full Text

R E T H I N K I N G H U M A N A T T E N T I O N A N D ITS C O M P O N E N T S by J E L E N A RISTIC B. A . , The University of British Co lumb ia , 2000 M. A . , The University of British Co lumbia , 2003 A T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y in T H E F A C U L T Y O F G R A D U A T E S T U D I E S (Psychology) T H E U N I V E R S I T Y O F BRIT ISH C O L U M B I A November 2006 © Je lena Rist ic, 2006. II Abstract Peop le attend to where others are looking. In three sect ions, spanning six studies and 11 experiments, the present thesis examines whether this socia l orienting effect is reflexive (Section I), if it should be cons idered a unique form of attentional orienting (Section II), and how it relates to traditional forms of reflexive and volitional spatial orienting (Sect ion III). The results from Sect ions I and II indicate that socia l attention can be conceptual ized as a strongly reflexive effect that is unique to eye gaze direction. However, other directional cues , like arrows, trigger similar spatial attention effects although their effects do not appear to be as strongly reflexive as the effects produced by eye gaze . The results from Sect ion III demonstrate that reflexive orienting elicited by an attentional cue posit ioned at fixation, such as gaze or an arrow, can occur independently of the orienting effects produced by cues that have typically been used to study reflexive and volitional orienting, i.e., nonpredictive peripheral onsets and predictive central cues, respectively. Taken together these results carry important implications for understanding socia l attention specif ical ly, and the conceptual izat ion and experimental examinat ion of human spatial attention in genera l . iii Tab le of Contents Abstract ii Table of Contents . iii List of Tab les v List of Figures vi Acknowledgements viii Co-Authorsh ip Statement ix C H A P T E R 1. Genera l Introduction 1 Modes of Attention Orienting 3 Dissertat ion Overv iew 8 References 10 S E C T I O N I: S O C I A L R E F L E X I V E A T T E N T I O N . . 12 Introduction. 12 References 15 C H A P T E R 2. Taking Control of Ref lexive Soc ia l Attention 16 Exper iment 1 18 Method 19 Resul ts 20 Discuss ion 21 Experiment 2 22 Method 23 Resul ts 23 Discuss ion 25 Genera l D iscuss ion 25 References 33 C H A P T E R 3. Eyes are Spec ia l but not for Everyone: The C a s e of Aut ism 35 Method 36 Resul ts 37 Discuss ion 38 References 42 S E C T I O N II: T H E U N I Q U E N E S S O F S O C I A L R E F L E X I V E A T T E N T I O N 44 Introduction 44 References 48 C H A P T E R 4. Are Eyes Spec ia l? It Depends on how you Look at it 49 Exper iment 1 52 Method 52 Resul ts 54 Discuss ion 55 Experiment 2 55 Method 56 Resul ts 57 Discuss ion 58 iv Experiment 3 58 Method 59 Resul ts 61 Discuss ion . 61 Genera l D iscuss ion 62 References 68 C H A P T E R 5. Attentional Effects of Counterpredict ive G a z e and Arrow C u e s 71 Experiment 1 76 Method 78 Resul ts 82 Discuss ion 83 Exper iment 2 87 Method 88 Resul ts 90 Discuss ion 91 Genera l D iscuss ion 93 References 108 C H A P T E R 6. Attentional Control and Ref lexive Orienting to G a z e and Arrow C u e s . . . 1 1 3 Experiment 1 117 Method 118 Resul ts 119 Discuss ion 120 Exper iment 2 121 Resul ts 121 Discuss ion 122 Genera l D iscuss ion 123 References 129 S E C T I O N III: I M P L I C A T I O N S F O R T R A D I T I O N A L M E A S U R E S O F A T T E N T I O N 132 Introduction 132 References 135 C H A P T E R 7. Nonpredict ive Central C u e s Trigger Independent Ref lexive Effects 136 Method 140 Resul ts 143 Discuss ion 149 References . 161 C H A P T E R 8. Genera l D iscuss ion 163 Reflexive Soc ia l Attention.. 166 Ref lexive Attention and Centra l C u e s 168 Implications for Understanding Ref lexive Orienting 188 Implications for Understanding Volit ional Orienting 190 Final Ref lect ions and Future Directions 193 Refe rences 198 A P P E N D I X I: U B C Behavioral Resea rch Ethics Board Certif icate of Approval 209 List of Tab les v C H A P T E R 2 Table 2.1. M e a n R T s , standard deviat ions, and error rates for Experiment 1 28 Table 2.2. Mean R T s , standard deviat ions, and error rates for Experiment 2 29 C H A P T E R 3 Table 3.1. Detai led Participant Information... 39 C H A P T E R 5 Table 5.1. Mean R T s (in ms), Standard Deviat ions, and Errors Rates (%) for Experiment 1 101 Table 5. 2. Mean R T s (in ms), Standard Deviat ions, and Errors Rates (%) for Exper iment 2 102 vi List of Figures C H A P T E R 1 Figure 1.1. A n Illustration, of the Posne r Cuing Task 9 C H A P T E R 2 Figure 2.1. Illustration of stimuli and sample sequence of events 30 Figure 2.2. Experiment 1 Resul ts 31 Figure 2.3. Experiment 2 Resul ts 32 C H A P T E R 3 Figure 3.1. Illustration of stimuli and sample sequence of events 40 Figure 3.2. M e a n response time (RT) in mi l l iseconds (ms) for high functioning individuals with autism (HFA) and typically developing (TD) individuals as a function of gaze cue - target delay when the target appeared at a location that was or was not gazed - at ...41 C H A P T E R 4 Figure 4 .1 . Illustration of stimuli and sample sequence of events for Exper iment 1 and Experiment 2 65 Figure 4. 2. Mean response time (RT) in mi l l iseconds (ms) for Experiment 1 (adult participants) and Exper iment 2 (child participants), as a function of cue-target stimulus onset asynchrony ( S O A ) and cue validity 66 Figure 4. 3. Illustration (not to scale) of stimuli, sample sequence of events, and mean response time (RT) for Experiment 3 (split-brain participant J .W.) , as a function of cue-target S O A , cue validity, and visual field 67 C H A P T E R 5 Figure 5.1. Illustration of the trial sequence in Exper iment 1 103 Figure 5. 2. Illustration of the three trial types that were possib le when gaze was directed at one of the four target locations in Exper iment 1... 104 Figure 5. 3. Experiment 1 mean R T s for counterpredictive gaze cues as a function of cue-target st imulus onset asynchrony ( S O A ) and trial type 105 Figure 5. 4. Illustration of the trial sequence in Exper iment 2 106 Figure 5. 5. Exper iment 2 mean R T s for counterpredictive arrow cues as a function of cue-target st imulus onset asynchrony ( S O A ) and trial type 107 C H A P T E R 6 Figure 6.1. Illustration of stimuli and sample sequence of events .127 Figure 6.2. Experiment 1 and Exper iment 2 results 128 C H A P T E R 7 Figure 7.1. Illustration of stimuli and sample sequence of events for the nonpredictive peripheral - nonpredictive arrow ( N P - N A ) cues condit ion 154 Figure 7.2. Illustration of stimuli and sample sequence of events for the predictive central - nonpredict ive arrow ( P C - N A ) cues condition 155 Figure 7.3. Illustration of stimuli and sample sequence for the nonpredictive peripheral -predictive central ( N P - P C ) cues condit ions 156 Figure 7.4. M e a n R T s for each individual cue type (NP, NA, P C ) as function of cue VII pairing ( N P - N A , P C - N A , N P - P C ) , cue validity and S O A 157 Figure 7.5. M e a n R T s for the nonpredictive peripheral - nonpredictive arrow (NP-NA) cue condition 158 Figure 7.6. M e a n R T s for the predictive central - nonpredictive arrow ( P C - N A ) cue condition 159 Figure 7.7. M e a n R T s for the nonpredictive peripheral - predictive central ( N P - P C ) cue condition 160 VIII Acknowledgements First, I would like to express s incere gratitude to my supervisor A lan Kingstone for giving me except ional gu idance and support during my graduate studies. I am truly privileged to have had you as an advisor. I would a lso like to acknowledge my co l leagues and col laborators, especial ly Chr is Fr iesen, for inspiring me through thought-provoking conversat ions and debates. Finally, the work presented here would not have been completed without the uncondit ional love, support, and encouragement that I have received from my parents Son ja and Pav le and especial ly from my husband S a s a . The research reported here was funded by junior and senior graduate fel lowships from the Natural Sc iences and Engineer ing Resea rch Counc i l of C a n a d a ( N S E R C ) and the Michael Smith Foundat ion for Health Research ( M S F H R ) , as well as by a Dissertation Resea rch Award from the Amer ican Psycholog ica l Assoc ia t ion (APA) . Addit ional support came from grants awarded to A lan Kingstone from the Human Frontiers Sc ience Program ( H F S P ) , N S E R C , and M S F H R . Co-Authorsh ip Statement I am the primary author on all the work presented in this P h D Dissertat ion, save for the study presented in Chapter 5, to which I have contributed equal ly as the first author. This study was included in the dissertation because its results are directly relevant to the main theme of the thesis. C H A P T E R 1 Genera l Introduction 2 In one of the earliest conceptual izat ions of human attention, Wil l iam J a m e s (1890) proposed that attention might be conceived as being either reflexive or volitional in nature. Accord ing to J a m e s , reflexive attention reflected pass ive activation by sensory events, whereas volitional attention reflected consc ious effort to actively attend to sensory events. This distinction between reflexive and voluntary attention has persisted into modern experimental psychology and helped to form one of the most influential behavioral tasks for studying human attention -- the Posne r cuing paradigm (e.g., Posner , 1978; Posner , 1980; Posner , Snyder , Dav idson, 1980). In this cuing paradigm, an abrupt peripheral onset stimulus, that does not predict spatial ly where a response target is likely to appear, is used to engage reflexive attention; and a central directional st imulus, usually an arrow, that does predict spatially where a response target is likely to appear, is used to engage volitional attention (e.g., Jon ides , 1981). Importantly, this c lear cut distinction between nonpredictive peripheral cues that isolate reflexive attention and predictive directional central cues that isolate volitional attention, has recently been brought into quest ion. Specif ical ly, there is now a substantial body of ev idence suggest ing that a picture of a face with the eyes directed toward a peripheral location will trigger a shift of attention to the cued (gazed-at) location even when the eyes are spatially nonpredict ive (e.g., Fr iesen & Kingstone, 1998; Driver et al . , 1999; Langton & Bruce, 1999). In other words, nonpredict ive central directional cues would appear to trigger reflexive attention. This P h D dissertation investigates the properties of the attentional shift elicited by eye direction and the implications of this result for attention research. In Sect ion I the ev idence that eyes as an attentional cue trigger a specia l form of reflexive socia l orienting is cons idered. In Sect ion II this social orienting effect for gaze direction is 3 considered relative to another familiar directional cue, the arrow. Finally, in Sect ion I l i a compar ison between reflexive orienting elicited by peripheral cues and reflexive orienting elicited by central cues is cons idered. Col lect ively, these data converge on the conclus ion that central cues activate reflexive orienting. The behavioral signature of this reflexive effect is similar for eye direction and arrow direction, although as suggested by studies in Sect ion I, and confirmed by studies in Sect ion II, eyes are more strongly reflexive than arrows thus making gaze direction unique. The general d iscuss ion considers the implications of these results for understanding both reflexive and volitional spatial orienting. Modes of Attention Orienting It is commonly conce ived by researchers in the field of human attention that spatial orienting can be al located in two qualitatively different ways : reflexively, somet imes cal led exogenous orienting; and volitionally, somet imes cal led endogenous orienting. Ref lexive shifts of attention are understood to occur when stimuli in the environment capture attention independent of one's goals and expectat ions. Volit ional shifts of attention, on the other hand, are understood to occur when attention is al located in accordance with one's internal goals and expectat ions. Both reflexive and volitional attention can be al located either in conjunction with eye movements , cal led overt orienting, or independently of any change in eye posit ion, cal led covert orienting (Klein, Kingstone & Pontefract, 1992; Klein & Shore , 2000; Klein, 2005). The exper iments presented in this dissertat ion have been des igned to measure covert attentional orienting, that is, orienting of attention across space that occurs when participants maintain central fixation and do not make eye movements. 4 This division of human spatial attention in terms of reflexive and volitional p rocesses is reflected in the prominent experimental cuing paradigm that is often referred to as the Posne r cuing paradigm (e.g., Posner , 1980; Posner , Snyder & Dav idson, 1980). This paradigm a s s u m e s that attention is a l imited-capacity resource which benefits the processing efficiency of information arriving at an attended location, resulting in lower response time (RT) and/or response errors. B e c a u s e human attention displays a capacity limitation, however, enhanced processing efficiency at an attended location is accompan ied by poorer processing efficiency at unattended locations, resulting in higher R T and/or response errors. Two distinct vers ions of the Posner cuing task have been deve loped, each speci f ic to the mode of attentional orienting to be measured. A peripheral cuing task (Posner, Snyder , Dav idson, 1980; Posner , 1980) was des igned to measure reflexive attentional orienting, and a central arrow cuing task (Posner, Snyder & Dav idson, 1980; Jon ides , 1981) was des igned to measure volitional attentional orienting. Ref lexive Attention Ref lexive attention has traditionally been conceptual ized as arising from the sensory pathway activation irrespective of an observer 's current goals and expectat ions. To maximize the observed benefits of reflexive spatial attention a suprathreshold st imulus, such as an abrupt luminance change, normally serves as the attentional cue (e.g., Posner , 1980). In the peripheral cuing task, the fundamental reflexive characterist ics of attentional orienting are considered to be revealed by requiring subjects to detect a target light at a peripheral location that was or was not preceded by an abrupt peripheral event, cal led the cue. Importantly, this peripheral cue does not predict where the target st imulus will appear. A s a result, any spatial effects of the cue on target detection are attributed to the reflexive orienting of attention toward the cued 5 location. The standard behavioral result is that R T to detect a target, typically measured by a s imple button-press response, is facilitated when a target appears at the cued location 300 ms or less after the cue was first presented. Beyond 300 ms, and lasting for up to a second or more, a target is detected more slowly at the cued location relative to a noncued location. This effect is cal led Inhibition of Return (IOR; Posner & C o h e n , 1984). The typical stimuli and sequence of events occurring in the peripheral cuing task are illustrated in Figure 1.1 A . Voluntary Attention In contrast to reflexive attention, voluntary orienting is conceptual ized as arising from consc ious al location of attentional resources by an observer towards an expected or current sensory event. To maximize the benefits of endogenous spatial attention, attentional cues manipulating voluntary orienting typically require development and maintenance of a spatial expectancy. In the central arrow cuing task, the fundamental volitional characterist ics of attentional orienting are thought to be revealed by requiring subjects to detect a target light at a peripheral location that was , or was not, pointed at by a central arrow. Importantly, the central arrow does predict where a target stimulus is likely to appear. B e c a u s e spatial effects of the central arrow are assumed to occur only when the arrow is spatially predictive, the observed attention effects are attributed to volitional orienting of spatial attention (Jonides, 1981). The standard result is that R T to the target at the cued location is facilitated for all cue-target intervals exceed ing 300ms with no ev idence of IOR (e.g., Klein, Kingstone & Pontefract, 1992). The typical stimuli and sequence of events occurr ing in the central arrow cuing task are illustrated in Figure 1.1B. 6 Spatial ly Nonpredict ive Central C u e s In contrast to the c lass ic conceptual izat ion of peripheral/reflexive and central/volit ional cuing descr ibed above, recent studies have reported a novel behavioral result suggest ing that spatially nonpredict ive cues presented at central fixation will trigger reflexive shifts of covert spatial attention. In this modified Posner cuing task, subjects are required to detect a target light appear ing at a location indicated by a central cue that does not predict where a target is going to appear. The typical results indicate that R T to respond to a target at the cued location is facilitated 100 ms after cue onset and this facilitation effect persists for cue-target de lays of 700-1000ms with no ev idence of IOR emerging (e.g., Fr iesen & Kingstone, 1998; Driver et al , 1999; Langton & Bruce, 1999). B e c a u s e the attentional effect in this paradigm is triggered by a directional cue that carries no reliable spatial information as to where the target is likely to appear, and because the behavioral effects of the cue emerge very rapidly, the observed orienting effect has been interpreted as reflecting a reflexive shift of attention to the cued location (e.g., Fr iesen & Kingstone, 1998; Langton & Bruce, 1999; Driver et a l . 1999). Figure 1.1C illustrates stimuli and a typical sequence of events used in the central nonpredict ive cuing task. In one of the first studies to employ spatially nonpredict ive central cues , Fr iesen and Kingstone (1998) presented a schemat ic face (the cue) on a computer screen that gazed to the left or right. Importantly, gaze direction did not predict where the target would appear. Fr iesen and Kingstone reported that within 100 ms of cue presentation, participants were faster to detect, local ize, and identify targets appear ing at a gazed-at target location compared to a not gazed-at target location. Driver et a l . (1999) and Langton and Bruce (1999) reported similar results using central uninformative eye direction cues (Driver et ai, 1999) and eye and head orientation cues (Langton and 7 Bruce, 1999) that were derived from photographs of real faces. Driver et a l . (1999) examined the influence of pupil deviation on a letter discrimination task and found that identification responses were a lways faster for targets appear ing at a gazed-at location. Similarly, Langton and Bruce (1999) examined the influence of pupil deviat ion and head orientation on target detection responses . Extending Fr iesen and Kingstone's (1998) and Driver et al 's (1999) data, their results indicated that head orientation a lso triggered reflexive shifts of spatial attention. Importantly, and in agreement with Fr iesen and Kingstone's (1998) results, the attentional effects obtained with pictures of real faces emerged early, 100-300 ms after the cue was presented, and persisted for approximately a second . Al though the interpretation that central nonpredictive cues trigger reflexive orienting of attention is consistent with the reported data, in their original study Fr iesen and Kingstone (1998) noted that in addition to displaying properties of reflexive orienting, the orienting effect shared two important properties with traditionally defined volitional orienting. Namely, consistent with a typical arrow cuing task, the orienting effect was triggered by a cue presented at central fixation (Jonides, 1981; Muller & Findlay, 1988; Mul ler & Rabbitt, 1989) and, the IOR effect, which is typically absent when observers orient attention voluntarily (e.g., Taylor & Kle in, 1998), was also absent when central spatially nonpredictive cues were used. B e c a u s e attentional orienting triggered by central eye gaze p o s s e s s e s both reflexive and volitional characterist ics, it has somet imes been conceptual ized as representing a "hybrid form of covert orienting" (Klein & Shore , 2000, p. 203) with "ambiguous" underlying control mechan isms (Klein, 2005) or more extremely, it has been speculated that orienting to central eye gaze is just an instance of top-down volitional orienting (e.g., Vece ra & R izzo , 2004; V e c e r a & R izzo , in press). 8 Dissertation Overv iew This P h D dissertat ion, which follows a manuscript based format, is compr ised of e leven exper iments that make up six studies. These studies are divided into three sect ions. In Sect ion I, two studies explore the proposed reflexive nature of the spatial attention effect triggered by central eye direction cues (Chapter 2 and 3). In Sect ion II, three studies investigate the supposed un iqueness of reflexive orienting to eye gaze relative to the familiar central arrow cue (Chapters 4-6). In Sect ion III, a novel cuing study compares reflexive orienting elicited by central and peripheral cues (Chapter 7). In the Genera l D iscuss ion (Chapter 8) the implications of the collective data for past and present conceptual izat ions of reflexive and volitional attention are d i scussed . F igure l .1 1.1 A : Peripheral Cuing 1.1B: Central Arrow Cuing 1.1C: Central G a z e Cu ing E Fixation Display C u e Onset Target Onse t • < < 5 0 % 5 0 % 8 0 % 2 0 % 5 0 % 5 0 % Figure 1.1. An Illustration of the Posner Cuing Task. Each row presents three stages (Fixation Display, Cue Onset, and Target Onset) that are standard to the paradigm. 1.1 A: Peripheral Cuing Task. At the start of each trial a central fixation dot is flanked by two squares. The left or right box is cued by brightening a box briefly (illustrated by the thick black line), and then a target is presented in either cued or uncued box. The task is to press a key as quickly as possible when the target is detected. The target appears in the cued box 50% of the time and in the uncued box 50% of the time. Thus, the cue does not predict where the target will appear. 1.1B: Central Arrow Cuing Task. The left or right box is cued by a central arrow pointing towards a box, and then a target is presented in either cued or uncued box. The task is to press a key as quickly as possible when the target is detected. The target appears in the cued box 80% of the time and in the uncued box 20% of the time. Thus, the cue predicts where the target will appear. 1.1C: Central Gaze Cuing Task. The left or right box is cued by eyes looking towards a box, and then a target is presented. The target appears in the cued box 50% of the time and in the uncued box 50% of the time. Thus, the cue does not predict where the target will appear. CD 10 References Driver, J . , Davis, G . , Ricciardel l i , P. , Kidd, P. , Maxwel l , E. , & Baron -Cohen , S . (1999). G a z e perception triggers visuospat ial orienting by adults in a reflexive manner. V isual Cogni t ion. 6, 509-540. Fr iesen, C . K., & Kingstone, A . (1998). The eyes have it!: Ref lexive orienting is triggered by nonpredict ive gaze . Psvchonomic Bulletin & Review. 5, 490-495. J a m e s , W . (1890) The Principles of Psycho logy. New York: Dover. Jon ides , J . (1981). Voluntary versus automatic control over the mind's eye 's movement. In J . B. Long and A . D. Badde ley (Eds.) , Attention & Per formance IX (pp. 187-203). Hi l lsdale, N J : Er lbaum. Klein, R. M., Kingstone, A . & Pontefract, A . (1992). Orienting of visual attention. In K. Rayner ( E d J , Eye Movements and Visual Cognition: Scene Perception and Reading, (pp. 46-63). North-Hol land: Elsevier Sc ience Publ ishers B. V . Kle in, R. M., & Shore , D. I. (2000). Relat ions among modes of v isual orienting. In S . Monsel l and J . Driver ( E d s J , Attention & Per formance XVIII (pp. 195-208). Cambr idge, M A : MIT P ress . Klein, R. M. (2005). O n the control of visual orienting. In M.I. Posner , (Ed.) Cognit ive Neurosc ience of Attention. New York: Guilford P ress . Langton, S . R. H., & Bruce, V . (1999). Ref lexive socia l orienting. V isua l Cogni t ion. 6, 541-567. Muller, H, J . & Findlay, J , M. (1988). The effect of visual attention on peripheral discrimination thresholds in single and multiple element d isplays. Ac ta Psvcho log ica . 69. 129-155. 11 Muller, H. J . , & Rabbitt, P. M. A . (1989). Ref lexive and voluntary orienting of visual attention: time course of activation and resistance to interruption. Journal of Exper imental Psycho logy : Human Percept ion and Per formance, 15, 315-330. Posner , M. (1978). Chronometr ic Explorat ions of Mind. Hi l lsdale, N J : Er lbaum. Posner , M. I. (1980). Orienting of attention. Quarterly Journal of Exper imental Psycho logy. 32, 3-25. Posner , M. I, Snyder , C . R. R, & Dav idson, B. J . (1980). Attention and detect ion of s ignals. Journal of Experimental Psycho logy: Genera l , 109, 160-174. Posner , M. I., & C o h e n , Y . (1984). Componen ts of V isua l Orienting. In H. B o u m a & D. G . Bowhuis (Eds.) , Attention & Per formance X (pp. 531-556). Hi l lsdale, N J : Er lbaum. Taylor, T. L , & Klein, R. M. (1998). On the causes and effects of inhibition of return. Psvchonomic Bulletin and Rev iew, 5, 6 2 5 - 6 4 3 . V e c e r a , S . P., & R izzo , M. (2004). What are you looking at? Impaired 'social attention' following frontal-lobe damage. Neuropsvcholog ia , 42, 1647 -1665 . S E C T I O N I: S O C I A L R E F L E X I V E A T T E N T I O N Introduction 13 In the first laboratory study to measure reflexive orienting to eye direction, Fr iesen and Kingstone (1998) asked participants to detect, local ize, or identify a peripheral target preceded by a central face that gazed left or right. Despite the fact that gaze direction and target location varied randomly from trial to trial, R T was shorter if a target appeared at a gazed-at versus a nongazed-at location. B e c a u s e this effect emerged early and occurred even though participants knew that gaze direction was irrelevant both to the task and location of the target, Fr iesen and Kingstone (1998) conc luded that gaze direction had triggered a reflexive shift in attention to the gazed-at location, possibly mediated by gaze-spec i f ic cel ls within the superior temporal su lcus (STS) . A similar position was advanced by Driver et al . (1999) and Langton and Bruce (1999). The first chapter of this thesis (Chapter 2) considers whether the reflexive gaze effect is sensit ive to top-down cognit ive control. To get at this issue, participants are shown an ambiguous st imulus cue that, depending on the top-down perceptual set adopted by participants, can be perceived either as a "car" or as containing "eyes" . Consis tent with a minor role for top-down cognit ive control, the data indicate that a gaze orienting effect emerges when participants are instructed to see the st imulus as having eyes . Importantly however, this reflexive orienting effect, once act ivated, cannot be abol ished by a change in the top-down perceptual set. Thus it appears that activation of the S T S may be sensit ive to top-down cortical input, but that once activated it cannot be inhibited easi ly (see Dolan et a l . , 1997; Bentin et al . , 2002 for similar conc lus ions from neuroimaging data). These first two experiments help to answer how attention comes to be oriented at a gazed-at location, but they do not answer why it is oriented there. This issue is considered in the final study of Sect ion I (Chapter 3). The existing literature suggests 14 two main explanat ions for why attention is shifted to a gazed-at location. One account, the "social reading hypothesis" is that attentional orienting occurs because eye direction is normally understood to be a meaningful socia l cue, for instance, communicat ing information between people about their internal states, e.g. , sexual attraction or disinterest (Baron-Cohen, 1995). The other explanat ion, the "feature cor respondence hypothesis" is that people are sensi t ive to the fact that features of the eyes are often pointed toward important events in the environment (Langton & Bruce, 1999). Chapter 3, reports an investigation that d issociates these often entangled accounts by examining social orienting for individuals with high functioning aut ism (HFA) and typically developing (TD) individuals. The data reveal that orienting to a gazed-at location normally occurs because of the socia l meaning that is read into the gaze direction of another individual. Together the results of these investigations indicate that orienting to where someone e lse is looking is a powerful, reflexive social phenomenon. Sect ion II considers whether this means that no other directional cue can have a similar effect, and if other cues can have a similar effect what this means for studies of reflexive orienting to gaze direction. 15 References Ba ron -Cohen , S . (1995). Mindbl indness: A n essay on autism and theory of Mind. Massachuse t ts : MIT Press . Bentin, S . , Sagiv , N., Meckl inger, A . , & Friederici , A . (2002). Pr iming visual face-processing mechan isms. Psycholog ica l Sc ience , 13(2), 190-193. Dolan, R., J . , Fink, G . R., Rol ls, E. , Booth, M., Holmes, A . , Frackowiak, R. S . J . , & Friston, K. J . (1997). How the brain learns to see objects and faces in an impover ished context. Nature. 389(9). 596-599. Driver, J . , Davis , G . , Ricciardel l i , P. , Kidd, P. , Maxwel l , E., & Baron -Cohen , S . (1999). G a z e perception triggers visuospat ial orienting by adults in a reflexive manner. V isua l Cogni t ion. 6 . 509-540. Fr iesen, C . K., & Kingstone, A . (1998). The eyes have it!: Ref lexive orienting is triggered by nonpredict ive gaze . Psychonomic Bulletin & Review, 490-495. Langton, S . R. H., & Bruce, V . (1999). Reflexive socia l orienting. V isua l Cogni t ion. 541-567. 16 C H A P T E R 2 Taking Control of Ref lexive Soc ia l Attention A v e r s i o n of this chapter has been publ ished. Ristic, J . & Kingstone, A . (2005). Taking control of reflexive socia l attention. Cognition, B55 -B65 . 17 Recent behavioral studies indicate that humans will attend to where someone is looking (e.g., Fr iesen & Kingstone, 1998; Langton & Bruce, 1999). In the typical laboratory demonstrat ion, a picture of a face looking left or right is projected onto a screen and observers are required to respond as quickly as possible to a target that appears beside the face. The standard finding is that response time (RT) to the target is shorter when the face is looking at the target rather than away from it, indicating that attention has been shifted to where the eyes are looking. Severa l reasons have been put forward for why this socia l orienting effect is reflexive in nature. First, it occurs rapidly, within a few hundred mi l l iseconds after a gaz ing face is presented (e.g., Fr iesen & Kingstone, 1998). S e c o n d , it occurs even if eye direction is negatively correlated with where a target might to appear (e.g., Driver et al , 1999; Fr iesen, Rist ic & Kingstone, 2004). Third, cel ls in the right inferior temporal (IT) cortex are dedicated to processing gaze direction in an obligatory fashion, which dovetai ls with the finding that attention is shifted rapidly to where someone e lse is looking (Langton, Watt & Bruce, 2000). Whether this socia l orienting effect is purely reflexive or not, however, has been the focus of considerable speculat ion. S o m e investigators have suggested, either explicitly or implicitly, that the effect is driven in a purely "bottom-up" fashion by cells in IT (e.g., Fr iesen & Kingstone, 1998; Kingstone, Fr iesen & G a z z a n i g a , 2000). For instance, in an early study of the socia l orienting phenomenon (Driver et a l , 1999) observers were shown gaz ing faces and informed that on most trials a target would appear at the location opposi te to where the eyes were looking (e.g., eyes looking left predicted that a target was likely to appear on the right). Even though the eyes were counterpredict ive, observers first shifted attention to the gazed-at location (where the 18 target was unlikely to appear), suggest ing that the initial attention shift triggered by gaze direction operates independent of top-down execut ive control p rocesses that are sensit ive to the predictive nature of a st imulus. There are, however, a lso reasons to think that the socia l orienting effect depends at least in part on top-down processes that interpret the trigger st imulus. For instance, Dolan et al (1997) observed that ambiguous pictures activated face-processing cel ls in IT only when observers recognized the pictures as depict ing faces. Similarly, Bentin and co l leagues (2002) have recently demonstrated that neutral stimuli, such as a pair of dots, will trigger a face-speci f ic brain potential only when the neutral stimuli are first represented as depicting the eyes of a face. Importantly, each of these lines of ev idence also has its shortcomings. For example, in the Driver et al . (1999) study observers never actually oriented attention to the predicted target location, raising the possibility that top-down control p rocesses were never even engaged . Converse ly , the studies by Dolan et al (1997) and Bentin et al (2002) lack behavioral data against which to compare the neural imaging results. Thus whether or not the socia l orienting effect, measured as a behavioral facilitation for targets appear ing at the gazed-at location, is driven purely by bottom-up processes remains very much an open quest ion. The aim of the present study was to address this issue directly. Experiment 1 The present study used ambiguous displays to a s s e s s whether top-down processes have a behavioral effect on attentional orienting to gaze direction. Part icipants were tested in one of three condit ions. In the F A C E condition (based on the original work of Fr iesen & Kingstone, 1998), participants were presented with a 19 schemat ic face that gazed to the left or right of center. Target onset occurred 100 -1000 ms after the face stimulus and was uncorrelated with gaze direction. In the other two condit ions participants were presented with an ambiguous stimulus (see Figure 2.1). In the E Y E S condition participants were instructed that the stimulus was a picture of a hat pulled down to the eyes of a face. In the C A R condition participants were instructed that the st imulus depicted an automobi le. Our predictions were as fol lows. In the basel ine F A C E condit ion we expected to replicate the results of Fr iesen and Kingstone (1998), and many others (e.g. Langton & Bruce, 1999; Rist ic, Fr iesen & Kingstone, 2003) with shorter R T at the gazed-at (valid) location versus the nongazed-at (invalid) location. A similar result was expected to emerge in the E Y E S condit ion, where the central st imulus would again be represented as gaz ing left or right. Two possible outcomes were plausible in the C A R condit ion. If face processing mechan isms in IT proceed in a purely modular bottom-up manner independent of top-down processing mechan isms, then performance in the C A R condit ion should replicate the E Y E S condit ion. That is, the cells in IT will ana lyze the stimulus as having the geometr ic shape of eyes , and trigger an attentional shift - a prediction well articulated by P i n k e r " . . . If objects other than faces (animals, facial express ions, or even cars) have some of these geometr ic features, the module will have no choice but to analyze them" (p 273). Alternatively, it is possible that top-down processing of the st imulus as depicting a gaz ing face is necessary for the socia l orienting effect to occur. If this is the case then in the C A R condit ion, and only in the car condit ion, a socia l orienting effect will be absent. Method Part icipants 20 All 45 participants were naive to the purpose of the experiment and ass igned randomly to one of the three groups (N=15/group). E a c h completed 10 practice trials fol lowed by 10 blocks of 60 trials for a total of 600 experimental trials. Ca tch trials, in which a target was not presented, varied randomly across trials and ranged from 6 -10% in a given block. In the F A C E condit ion, participants were informed that the central st imulus depicted a face, and that eye direction did not predict target posit ion. The instructions for the E Y E S and C A R condit ions were carefully scripted so that the only difference between the two was the information regarding the identity of the central st imulus, i.e., a hat pulled down to the eyes or a car. Part icipants were informed that any changes in the central st imulus (e.g., eyes or car) did not predict target posit ion. Resul ts Key press errors, false alarms, anticipations (RT< 100 ms), and slow R T s (RT> 1000 ms) were classif ied as errors and excluded from analys is. For all condit ions, false alarms occurred on less than 4 . 3 3 % on catch trials. Addit ionally, less than 2 .6% of all target present trials in each cue condition were tr immed because of errors. Mean RT, standard deviat ions, and error rates for each condition are presented in Table 2.1. M e a n R T s were calculated for correct target trials for each condit ion as a function of validity and S O A across all participants. The means are illustrated in Figure 2. 2 and show that for both the F A C E and E Y E S condit ions R T was shorter when a target appeared at a gazed-at (valid) versus a nongazed-at (invalid) location, i.e., the socia l attention effect. In contrast, there was no reliable effect of validity in the C A R condit ion. 21 T h e s e observat ions were supported by a 3 x 2 x 4 analysis of var iance ( A N O V A ) with condition ( F A C E , E Y E S , C A R ) as a between subject factor and validity (valid, invalid) and S O A (100, 300, 600, and 1000 ms) as within subject factors. There were main effects for validity [F(1,42)=19.97,p<.0001] reflecting the overall facilitative effect of attention being al located to a valid location; and S O A [F(3,126)=91.69,p< 0001] reflecting the general decl ine in R T that occurs as participants prepare to respond to a target (called a foreperiod effect; Berte lson, 1967). S O A also interacted with condit ion [F (6,126)=2.48,p<0.05], and validity, [F (3,126)=6.43,p< 0.001] reflecting that the foreperiod effect was most pronounced in the F A C E and E Y E S condit ion, and when the target was at the valid location. Most importantly, there was a significant condit ion x validity interaction [F (2,42)=3.41,p< 0.05] consistent with attention being al located to the valid location in the F A C E and E Y E S condit ions but not in the C A R condit ion. In agreement with this interpretation, when each condit ion is ana lyzed individually, there is a significant main effect of validity for the F A C E and E Y E S condit ions [both Fs>9.4,ps<. 01] but not for the C A R condition [F<1; the only significant effect being S O A [F (3,42)=19.98,p< 0001]. D iscuss ion The results of Exper iment 1 were c lear cut. Attention was shifted reflexively by stimuli that were represented as eyes in the F A C E and E Y E S condit ions. However, the very same ambiguous stimulus used in the E Y E S condit ion failed to trigger reflexive orienting in the C A R condit ion. A s noted in the introduction to Experiment 1, this data pattern agrees with the position that bottom-up orienting mechan isms triggered by perceived gaze direction are modulated by top-down p rocesses . W e return to this issue in the general d iscuss ion. 22 The reason why we chose to ass ign different participants to different condit ions in Exper iment 1 was because there is recent neuroimaging ev idence suggest ing that once people perceive an ambiguous stimulus as representing a face, they have difficulty representing it as another type of object (Bentin & Gol land, 2002). In Exper iment 2 we turned this bias toward face representation to our advantage. Al l the participants in Exper iment 2 received both the E Y E S condition and the C A R condit ion, with half receiving the E Y E S condit ion first and half receiving the C A R condit ion first. Experiment 2 Manipulat ing the E Y E S and C A R condit ions within the same participants is crucial for two reasons. First, a between group compar ison of performance after the first half of testing provides a direct replication of the E Y E S versus C A R compar ison in Experiment 1. Here we expected that if the difference we observed previously between these condit ions is real and replicable we should find again that attention is shifted only in the E Y E S condit ion. S e c o n d , and most importantly, a different prediction is made for the second half of testing. Here we expected that the participants who had first received the C A R condition would now show ev idence of reflexive orienting in the E Y E S condit ion because the central st imulus would now be perceived as a face. This prediction stands in contrast to the outcome expected for the participants who had received the E Y E S condition first. Because of the asymmetry noted above, where a st imulus persists in being perceived as a face once it is seen as a face, we expected that participants who received the C A R condit ion second - that is, after receiving the E Y E S condition - would continue to show a validity effect in that condit ion. 23 Method All 36 participants were naive to the purpose of the experiment and to the condition change that occurred halfway through testing. The apparatus and the ambiguous fixation stimuli were the s a m e as in Exper iment 1. Des ign and procedure were also the same with the following except ions. Half the participants received the E Y E S condition before the C A R condit ion; the remaining participants received the reverse order of condit ions. E a c h condit ion was preceded by 10 practice trials fol lowed by 8 blocks of 60 trials, for a total of 960 test trials. Instructions for these condit ions were as before. Resul ts Fa lse alarms occurred on less than 2 .5% of the catch trials, and less than 0.5% of all target present trials were in error. Mean R T s , standard deviat ions, and their assoc ia ted error rates are presented in Table 2. M e a n R T s for correct target trials were calculated for each participant. Interparticipant means across S O A and validity condit ions for both condit ions are shown in Figure 2. 3. To test whether the effects observed in Experiment 1 were replicated in the present study, we conducted a 2 (condition) x 2 (validity) x 4 ( S O A ) A N O V A with E Y E S [first] versus C A R [first] as a between subject factor and validity and S O A as within subjects factors. The results replicated Experiment 1, with significant main effects of validity [F(1,34)=16.57,p< 0005] and S O A [F(3,102)=85.24,p<0001] as well as the crucial interaction between condit ion and validity [F(1,34)=4.26,p<.05] reflecting again the presence of a validity effect in the E Y E S condition and the absence of one in the C A R condit ion. No other effects were significant. 24 W e had predicted that both the E Y E S [second] and C A R [second] condit ions would reveal a significant effect of validity. A 2 (condition) x 2 (validity) x 4 ( S O A ) A N O V A confirmed this prediction. The main effects of validity [F (1,34)=18.9,p<.0001] and S O A [F(3,102)=68.98,p<0001] were highly significant with no significant interactions (all Fs<1.8, ps>.14). In particular, there was no condit ion x validity interaction (F<1), demonstrat ing that there was a significant, and equivalent, effect of validity for both the E Y E S and C A R condi t ions 1 . Together these data converge on the conclus ion that the validity effect varied as a function of condit ion only for those participants that received the C A R condition first. This was confirmed by two separate within-group 2 (condition) x 2 (validity) x 4 ( S O A ) A N O V A s . For the C A R [first] - E Y E S [second] group, main effects of validity [F (1,17)=9.28,p< 01] and S O A [F (3,51 )=55.54,p<0001] were highly significant, as was the condition x validity interaction [F(1,17)=4.61 ,p<.05]. No other effects were reliable [all Fs<2.1, all ps>.1]. In contrast, for the E Y E S [first] - C A R [second] group, the main 1 Note that, as in Experiment 1, when C A R [first] was analyzed using a separate 2 (validity) x 4 (SOA) within-subjects ANOVA, only a main effect of S O A was significant [F (3,52)=35.71 ,p< 0001]. The lack of a significant validity effect (p>.12) or validity x S O A interaction (p>.19) indicates there was no social attention effect in this condition. 2 We were concerned that the validity effect emerged in the C A R [first] - E Y E S [second] group because of practice effects rather than because of the perception of the ambiguous stimulus as possessing eyes. A close examination of the data eliminated this concern. For Experiment 2 we compared the last two blocks (blocks 7 and 8) of the first condition with the first two blocks (blocks 9 and 10) of the second condition. As before, in the C A R [first] - E Y E S [second] group there was a significant condition x validity interaction [F (1,17)=4.6,p<. 05] reflecting the emergence of a validity effect when the condition was switched from C A R to E Y E S . In contrast, in the E Y E S [first] - C A R [second] group a significant validity effect was observed [F (1,17)=6. 55,p<. 05] which persisted across conditions [condition x validity interaction, F<1]. Critically, when the C A R condition in Experiment 1 was examined in an identical manner (blocks 7 and 8 vs. blocks 9 and 10), there were no significant effects involving validity (all Fs<1). Together these data demonstrate conclusively that the validity effect emerged in the C A R [first] - E Y E S [second] group, and persisted in the E Y E S [first] - C A R [second] group, because of the perception of the ambiguous stimulus as possessing eyes. 25 effects of validity [F (1,17)=23.52,p< 001] and S O A [F (3,51)=46.62,p< 0001] but importantly there was no interaction involving validity (all ps >.17) 2 . D iscuss ion The results of Exper iment 2 extended the results reported in Exper iment 1 in two important ways . O n e , we found again that attention was shifted reflexively when the ambiguous stimulus w a s first perceived as E Y E S but not when it was initially seen as a C A R . Importantly when these participants in the C A R condit ion were presented with the E Y E S condit ion, they began to shift attention reflexively. Two, we found that the participants who received the E Y E S condit ion first continued to shift attention reflexively when presented with the C A R condit ion. This new result converges with, and provides behavioral support, for the Bentin and Gol land (2002) finding that once an ambiguous stimulus is perceived as a face it will persist in being perceived as such . Genera l D iscuss ion Attention is shifted reflexively to where someone else is looking. A wealth of ev idence implicates face processing mechan isms specif ic to inferior temporal (IT) cortex as being crucial to this social attention effect (e.g., Kanwisher , 2000; Hoffman & Haxby, 2000). In the present study we asked whether this socia l attention effect is driven by neurons in IT in a purely bottom-up manner independent of top-down control p rocesses responsib le for stimulus interpretation. The answer is no. The reflexive social attention effect is modulated by top-down control p rocesses . Two lines of ev idence in the present study converge on this conclus ion. First, in Experiment 1, we showed that whether the same stimulus triggers a reflexive shift in attention depends on how it is perceived by the observer. That the 26 absence or presence of the reflexive socia l attention effect can be triggered by a slight change in st imulus interpretation demonstrates that this attention effect is sensit ive to top-down control. S e c o n d , in Exper iment 2, we found an asymmetry in the ability to manipulate the attention shift triggered by the ambiguous stimulus. Specif ical ly, when first informed that the stimulus was a C A R and then later informed that it contained E Y E S , an attention shift was observed only in the E Y E S condit ion. However, when first informed that the st imulus possessed E Y E S , and then that it was a C A R , the attention shift in the E Y E S condition persisted into the C A R condit ion. This provides strong and convergent behavioral ev idence that once top-down processes lead to the perception of a stimulus as a face, it is extremely difficult to avoid see ing that st imulus as a face. Together the data go a long way toward reconcil ing a point of contention within the field - whether or not the reflexive socia l attention effect is sensit ive or not to top-down control. O n the one hand our study shows clearly that the socia l attention effect is sensit ive to top-down control insofar as determining whether a st imulus is at first perceived as possess ing facial features or not. O n the other hand, the socia l attention effect is not sensit ive to top-down control insofar as a stimulus will persist in being seen as having face features once it has been perceived that way. This latter finding highlights why the socia l attention effect must ultimately be considered as reflexive in nature, for once a st imulus activates IT and is perceived as having features such as eyes , the attentional effect of this stimulus appears to be insensit ive to top-down modulat ion. This complex interplay between reflexive and volitional attention, and how the activation of bottom-up p rocesses may rely on execut ive top-down processes , dovetai ls with a growing recognition that reflexive attention may depend ultimately on 27 the meaning that individuals attach to stimuli (see Rauschenberger , 2003 for a recent review on this issue). 28 Table 2.1. Condit ion F A C E E Y E S C A R 100 ms S O A Val id Invalid M 341 342 S D % E 49 50 .5 .8 M 353 354 S D % E 55 57 .2 .8 M 350 349 S D % E 36 36 0 .3 300 ms S O A Val id 318 38 2.3 322 53 1.6 322 42 1.3 Invalid 329 43 2.3 337 53 2.2 328 43 1.3 600 ms S O A Val id 297 38 .8 305 44 .4 312 37 .5 Invalid 309 42 .7 318 46 .4 314 37 .6 1000 ms S O A Val id 309 46 .6 314 39 .7 332 41 .9 Invalid 313 41 .6 321 47 .3 329 37 .5 Table 2.1. Mean R T s , standard deviat ions, and error rates for Exper iment 1. 29 Table 2.2. G R O U P Condit ion C A R C A R - E Y E S E Y E S E Y E S - C A R E Y E S C A R M S D % E 100 ms S O A Val id 336 33 .1 Invalid 339 33 .1 300 ms S O A Val id 307 29 2 Invalid 315 33 1.5 600 ms S O A Val id 306 35 .2 Invalid 305 27 .8 1000 ms S O A Val id 317 34 .2 Invalid 315 32 .2 M 335 341 305 315 302 307 311 314 S D % E 36 43 34 42 38 36 36 32 .3 .1 2.1 1.4 .3 .5 .4 .3 M S D % E 330 35 337 34 304 313 288 298 33 30 29 28 305 31 307 31 .3 .2 .7 .8 .3 .4 .3 .7 M S D % E 323 39 325 42 305 43 308 38 287 33 296 37 .9 .2 .9 1.5 .7 1.0 299 37 .6 301 30 .2 Table 2.2. M e a n RT , standard deviat ions, and error rates for Exper iment 2. 30 Figure 2.1 FACE E Y E S OR CAR Figure 2.1. Illustration of stimuli (not to scale) and sample sequence of events. Every trial began with a 675 ms presentation of a fixation point (subtending 1°) followed by a central stimulus cue (FACE, E Y E S , or CAR) . The stimulus onset asynchrony (SOA) separating the presentation of the central cue and the target was 100, 300, 600, or 1000 ms. Cue direction (e.g., eyes left or right), target position, and S O A were varied randomly. Participants were instructed to maintain central fixation and press the spacebar on a computer keyboard as quickly as they could when the target was detected. Both the central stimulus and the target remained present until a response was made or the trial timed-out after 2700 ms. Response time (RT) was measured from the onset of the target. The intertrial interval was 680 ms. The central stimulus condition was manipulated between participants. All computer stimuli were black drawings shown on a white background. The F A C E stimulus was comprised of a circle outline (8.2° long and 7.2° wide) with two inner upper circles representing eyes, middle small circle representing the nose (0.2°) and the straight line representing the mouth (2.5° in length). The circle outline of eyes subtended 1 and filled-in circles representing pupils measured 0.6°. The pupils were positioned so that they were either touching the left or right circle outline. The central stimulus was identical for the E Y E S and C A R conditions. This stimulus was a symmetrical black and white line drawing. It measured 5° in width and 4° in height. The line drawings of three circles subtended 1° and black filled in circles measured 0.6°. The target was a black asterisk appearing on either left or right side of the central cue with an eccentricity of 7°of visual angle. The asterisk was 1° high and 0.9°wide. Figure 2.2 Figure 2. 2. Experiment 1 results. Figure 2.2 shows mean RTs in milliseconds as a function of S O A and validity for the three stimulus cue conditions (FACE, E Y E S , CAR) manipulated in Experiment 1. Error bars depict standard error of the difference of the means. 32 280 100 ms 300 ms 600 ms 1000 ms Stimulus Onset Asynchrony (SOA) 100 ms 300 ms 600 ms 1000 ms Stimulus Onset Asynchrony (SOA) E Y E S C A R 360 £ . 344 a) 3 2 8 h l -g ro 0 01 ro 312 296 h 280 100 ms 300 ms 600 ms 1000 ms 100 ms 300 ms 600 ms 1000 ms Stimulus Onset Asynchrony (SOA) Stimulus Onset Asynchrony (SOA) Figure 2.3. Experiment 2 results. Figure 2.3 illustrates mean RTs in milliseconds as a function of stimulus onset asynchrony (SOA) and stimulus cue validity in Experiment 2. The top row illustrates results for the C A R [first]-EYES [second] group and the bottom row shows the results for the E Y E S [first]-CAR [second] group. Error bars depict standard error of the difference of the means. 33 References Bentin, S . , Sagiv , N., Meckl inger, A . , & Friederici, A . (2002). Pr iming visual face-processing mechan isms. Psy j ^hp jo j j i ^ ^ Bentin, S . & Gol land, Y . (2002). Meaningful process ing of meaning less stimuli: The influence of perceptual exper ience on early visual process ing of faces. Cognit ion, 86, B1 -B14 . Berte lson, P. (1967). The time course of preparation. Quarterly Journal of Exper imental Psycho logy. 19, 272-279. Dolan, R., J . , Fink, G . R., Rol ls, E. , Booth, M. , Holmes, A . , Frackowiak, R. S . J . , & Friston, K. J . (1997). How the brain learns to see objects and faces in an impover ished context. Nature. 389(9), 596-599. Driver, J . , Davis, G . , Ricciardel l i , P., Kidd, P., Maxwel l , E., & Baron -Cohen , S . (1999). G a z e perception triggers visuospat ial orienting by adults in a reflexive manner. V isua l Cogni t ion, 6 , 509-540. Fr iesen, C . K., & Kingstone, A . (1998). The eyes have it!: Ref lexive orienting is triggered by nonpredictive gaze . Psychonomic Bulletin and Rev iew, 5, 490-495. Fr iesen, C . K., Rist ic, J . & Kingstone, A . (2004). Attentional effects of counterpredictive gaze and arrow cues . Journal of Exper imental Psycho logy: Human Percept ion and Per formance. 30. 319-329. Hoffman, E. A . & Haxby, J . V . (2000). Distinct representat ions of eye gaze and identity in the distributed human neural sys tem for face perception. Nature Neurosc ience, 3, 80 - 84. Kanwisher , N. (2000). Domain-specif ic i ty in face perception. Nature Neurosc ience, 3(8), 759-763. Kingstone, A . , Fr iesen, C . K., & G a z z a n i g a , M. S . (2000). Ref lexive joint attention depends on lateralized cortical connect ions. Psycholog ica l Sc ience . 11. 159-165. Langton, S . R. H., & Bruce, V . (1999). Ref lexive socia l orienting. V isua l Cognit ion, 6, 541-567. Langton, S . R. H., Watt, R. J . , & Bruce, V . (2000). Do the eyes have it? C u e s to the direction of social attention. Trends in Cognit ive Sc iences , 4, 50-59. Pinker, S . (1997). How the Mind Works . New York, N Y : Norton. Rauschenberger , R. (2003). Attentional capture by auto- and al lo-cues. Psvchonomic Bulletin and Rev iew, 10. 814-842. Ristic, J . , Fr iesen, C . K., & Kingstone, A . (2002). A re eyes spec ia l? It depends on how you look at it. Psycholog ica l Bulletin and Rev iew, 9, 507-513. C H A P T E R 3 Eyes are Spec ia l but not for Everyone: The C a s e of Aut ism A version of this chapter has been publ ished. Rist ic, J . , Mottron, L , Fr iesen, C . K., larocci, G . , Burack, J . , & Kingstone, A . (2005). Eyes are spec ia l but not for everyone The case of aut ism. Cognitive Brain Research, 24, 715-718. 36 W h y do we have a tendency to shift our attention to where other people are looking? Investigations suggest that there are two possible explanat ions. One is that eye direction conveys key socia l information, such as status, personal interest, and attentional engagement (Baron-Cohen, 1995). W e call this the socia l reading hypothesis. The other is that people are sensit ive to changes in the basic st imulus features that are assoc ia ted with shifts in gaze direction, in particular, the cor respondence between the location of an interesting event in the environment and the position of the pupils/ ir ises in the eyes that are directed towards that location (Langton & Bruce, 1999; Fr iesen & Kingstone, 1998). W e call this the feature cor respondence hypothesis. In the past, these two conceptual izat ions were tied s o c losely to one another that they were often d i scussed as though they were synonymous, as it is difficult to imagine a natural situation in which the social meaning assoc ia ted with gaze direction and the perceptual features assoc ia ted with gaze direction could be disentangled (Driver et al , 1999). In the present study, we show that the two indeed can be d issoc iated, a finding that carries substantial implications for the understanding of human social cognit ion. Method W e examined the performance of a total of 47 participants (see Table 3.1), who v iewed static d isplays of left- and right-deviated gaze on a computer sc reen. The participants were asked to make a speeded keypress response when they detected a target occurr ing to the left or right of the face following one of four gaze cue - target delay intervals (see Figure 3.1). Both high functioning individuals with aut ism (HFA) and typically developing peers (TD) were ass igned randomly to either the nonpredictive gaze condit ion or the predictive gaze condit ion. In the nonpredict ive condit ion, a target appeared at the gazed-at location 5 0 % of the time and at the not-gazed-at location 5 0 % 37 of the time. In the predictive condit ion, a target appeared at the gazed-at location 8 0 % of the time and at the not-gazed-at location 2 0 % of the time. Resul ts W h e n eye direction was spatially predictive, as illustrated in F igures 3.2A & 3 .2C, both H F A s and T D s were faster to detect targets occurring at the gazed-at location. This indicates that both groups could perceive and use gaze direction as an attentional cue when the cor respondence information was known to be a reliable aid to the task at hand. The key quest ion, however, is whether attending to a gazed-at location is driven by the stimulus features (high cor respondence between the cue and the target) or by social re levance of perceived gaze direction. Figures 3.1B and 3.1D show that, consistent with previous findings (Fr iesen & Kingstone, 1998) T D s shifted their attention in response to perceived eye direction when it was spatially nonpredict ive (that is, when the cor respondence between eye direction and target location was at chance) . In contrast, H F A s did not shift their attention in response to nonpredictive eye direction. This difference in performance supports the socia l reading hypothesis as T D s , who can respond to the socia l power of eyes , orient automatically in response to gaze direction even when it conveys no predictive information about environmental events. In contrast, H F A s do not attend to eye direction when it is spatially nonpredictive. This is consistent with the notion that H F A s are not sensit ive to eyes as displaying social ly relevant information (Grelotti, Gauthier & Schul tz , 2002; Klin et a l , 2002; Shul tz et al , 2000) but are exquisitely sensit ive to changes in event probability in their environment (Klinger & Dawson , 2001). Thus , our experiment supports the feature cor respondence hypothesis with regard to H F A s . It a lso highlights the outcome that H F A s essential ly outperform the T D s in the nonpredict ive condit ion insofar as H F A s were not " fooled" by a nonpredict ive gaze cue. 38 Discuss ion The results of the present study suggest that efforts to train individuals with aut ism to use eye direction as a probabilistic feature cor respondence cue fail to capture the key and fundamental component that gaze direction is normally used as a cue that is prioritized by the human attention system because of its social re levance. At best, individuals with autism appear to learn to orient attention to features that are usually confounded with eye direction, such as abrupt transients and stimulus motion (Swettenham et al , 2003; Chawarska , Klin & Volkmar, 2003; Kyl l iainen & Hietanen, 2004; Senju et a l , 2004). This failure to appreciate the socia l power of human eyes appears to be grounded in fundamental dif ferences in brain function between individuals with autism and the general population. For example, functional neuroimaging ev idence indicates that the superior temporal su lcus ' typical special izat ion for process ing faces and eyes (Al l ison, P u c e & McCar thy , 2000; Hoffman & Haxby, 2000; Deaner & Piatt, 2003) is not evident consistently in individuals with autism (Klin et al , 2002; Schul tz et a l , 2000). Our study provides ev idence that perceived socia l re levance, and not feature cor respondence, drives automatic attentional orienting in response to gaze direction for typically developing individuals, but that feature cor respondence, and not socia l re levance, mediates attention to gaze effects in individuals with aut ism. A s such , the present study provides the first dissociat ion of these two often-confounded explanat ions, carrying with it important implications for understanding the development of social attention in both healthy and atypical populations. 39 Table 3.1. Condit ion Group N M e a n IQ M e a n C A Predict ive G a z e H F A 12(11 males) 106.1 17.1 T D 11 (7 males) 100.1 15.0 Nonpredict ive G a z e H F A 12 (12 males) 110.8 20.5 T D 12 (12 males) 114.8 21.8 Table 3.1. Detailed participant information. Al l individuals included in High Funct ioning Aut ism (HFA) group met the diagnost ic criteria for H F A or Asperger Syndrome according to the Aut ism Diagnost ic and Observat ional Schedu le -Gener i c (Lord et a l , 2000) and the Aut ism Diagnost ic Interview-Revised (Lord et a l , 1994). Part icipants included in the typically developing (TD) group were sc reened for history of psychiatric disorders. Groups were matched for mean IQ and chronological years of age (CA) . Four participants (not shown) were exc luded due to failures to perform the speeded aspect of the task. Figure 3.1. 40 y time >-Fixation Display G a z e C u e Onset Target Onset 675 ms 105,300,600, until response, or 1005 ms or 2700 ms Figure 3.1. Illustration (not to scale) of stimuli and sample sequence of events. The start of each trial was signaled by the presentation of a schematic face with blank eyes. The pupils, looking left or right, appeared 675 ms later. A target presented at an eccentricity of 3.55° of visual angle was shown on the left or the right side of the fixation stimulus 105, 300, 600, or 1,005 ms after appearance of the pupils (the attentional cue). Both the central face and the target remained on the screen until a response was made or 2, 700 ms had elapsed, whichever came first. Speeded response time was measured from the onset of the target. Participants were instructed to maintain central fixation and were informally monitored as it is well established that eye movements do not occur when suprathreshold targets such as those used here must be detected. In the nonpredictive cue condition each participant received a total of 336 experimental trials, divided equally over 8 testing blocks while in the predictive cue condition each participant received 672 trials, divided equally over 16 testing blocks. 41 Figure 3.2 High Functioning Autism Figure 2A: Predictive Gaze 450 424 k 346 k 320 gazed-at not gazed-afj 450 424 «T 398 E H 372 346 320 Figure 2B: Nonpredictive Gaze —B— gazed-at - • - not gazed-at 450 424 | 398 r2 372 346 320 450 Typically Developing Figure 2C: Predictive Gaze —•— gazed-at • - Q - not gazed-at -• s . LD • 1 1 1 1 Figure 2D: Nonpredictive Gaze 424 h l" 398 c 372 346 320 gazed-at not gazed-atl 100 ms 300 ms 600 ms 1000 ms Cue-Target Delay 100 ms 300 ms 600 ms 1000 ms Cue-Target Delay Figure 3.2. Mean response time (RT) in milliseconds (ms) for high functioning individuals with autism (HFA) and typically developing (TD) individuals as a function of gaze cue - target delay when the target appeared at a location that was or was not gazed-at. The column on the left shows HFA performance when gaze direction was predictive (top figure) and when it was nonpredictive (bottom figure) of target location. The column on the right shows TD performance when gaze was predictive (top figure) and when it was nonpredictive (bottom figure) of target location. Mean response error never exceeded 3.6% and averaged .9%. An analysis of variance (ANOVA) of HFA RT performance in the predictive condition, with gaze-target stimulus onset asynchrony (SOA) and gaze-target validity as factors, revealed significant main effects for both S O A and gaze validity [both Fs>8.5, p's<. 005] with the validity effect growing as a function of S O A [F (3,33)=4.52,p<. 01]. In the nonpredictive condition there was no effect for gaze direction and no interaction between gaze and S O A [both Fs<1] although there was a main effect of S O A [F>35, p<. 01]. A between-subjects ANOVA comparing the two conditions confirmed that there was a significant gaze x predictiveness interaction [F (1,22)=11.23, p<. 01]. The same set of analyses conducted on the TD data revealed that for the predictive condition, and for the nonpredictive condition, all main effects were significant [all Fs>10.4 p<. 01] as were the gaze x S O A interactions [both Fs>2.93,p<. 01]. A between-subjects A N O V A comparing the two conditions revealed that the gaze x predictiveness interaction was not significant [F(1,21)=1.0,p>. 3]. 42 References Al l ison, T., P u c e , A . & McCar thy G . (2000). Soc ia l perception from visual cues : role of S T S region. Trends in Cognit ive Sc ience 4, 267-278. Ba ron -Cohen , S . (1995). Mindbl indness: A n essay on aut ism and theory of Mind. Massachuse t ts : MIT P ress . Chawarska , K., Kl in, A . & Volkmar F. (2003). Automat ic attention cueing through eye movement in 2-year old children with aut ism. Chi ld Development 74, 1108-1122. Deaner, R. O. , & Piatt, M. L. (2003). Ref lexive socia l attention in monkeys and humans, Current Biology. 13, 1609-1613. Driver, J . , Davis, G . , Ricciardel l i , P. , Kidd, P. , Maxwel l , E., & Baron -Cohen , S . (1999). G a z e perception triggers visuospat ial orienting by adults in a reflexive manner. V isual Cognit ion. 6, 509-540. Fr iesen, C . K., & Kingstone, A . (1998). The eyes have it!: Ref lexive orienting is triggered by nonpredictive gaze . Psychonomic Bulletin & Review. 5, 490-495. Grelotti, D. J . , Gauthier, I. & Schul tz R. (2002). Soc ia l interest and the development of cortical face special izat ion: What aut ism teaches us about face process ing. Developmental Psychob io logy 40. 13-25. Hoffman, E. , A . , & Haxby, J . , V . (2000). Distinct representations of eye gaze and identity in the distributed human neural sys tem for face perception. Nature Neurosc ience, 3, 80-84. Kl in, A . , Jones , W. , Schul tz , R., Vo lkmar F. & C o h e n D. (2002). V isua l fixation patterns during viewing of naturalistic socia l situations as predictors of social competence in individuals with aut ism. Arch ives of G e n Psychiatry. 59 809-816. Kl inger L . G . & Dawson G . (2001). Prototype formation in aut ism. Developmenta l Psvchopathology, 13, 111-24. 43 Kyll iainen A . & Hietanen J . K. (2004). Attention orienting by another's gaze direction in children with aut ism. Journal of Chi ld Psycho logy and Psychiatry. 45, 435-444. Langton, S . R. H., & Bruce, V . (1999). Ref lexive socia l orienting. V isua l Cognit ion. 6, 541-567. Lord, C , et al . (2000). The Aut ism Diagnost ic Observat ions Schedu le - Gener ic : A standard measure of socia l and communicat ion deficits assoc ia ted with the spectrum of aut ism. Journal of Aut ism and Developmenta l Disorders. 30, 205 -223. Lord, C , Rutter M. & Le Couteur A . (1994). Aut ism Diagnost ic Interview-Revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. Journal of Aut ism and Developmental Disorders. 24, 659-685. Senju, A . , Tojo, Y . , Dairoku, H. & H a s e g a w a T. (2004). Ref lexive orienting in response to eye gaze and an arrow in children with and without aut ism. Journal of Chi ld Psycho logy and Psychiatry, 45 , 445-458. Schul tz R.T. et a l . (2000). Abnormal ventral temporal cortical activity during face discrimination among individuals with autism and Asperger 's syndrome. Arch ives of Genera l Psychiatry, 57, 331-340. Swet tenham, J . , Cond ie , A . , Campbe l l , R., Milne, E. & Co leman , M. (2003). Does the perception of moving eyes trigger reflexive visual orienting in aut ism? Phi losophical Transact ions of Roya l Society of London: B Biolological S c i e n c e s 358, 325-334. S E C T I O N II: T H E U N I Q U E N E S S O F S O C I A L R E F L E X I V E A T T E N T I O N Introduction 45 Fr iesen and Kingstone (1998) argued that reflexive orienting triggered by eye gaze direction is reflexive, that it represents an attentional p rocess that is unique to biologically relevant stimuli, and that it may be subserved by brain regions specif ic to the processing of eyes , i.e., the S T S . The two studies presented in Sect ion I support the notion that orienting to gaze direction represents a reflexive socia l attention phenomenon. Specif ical ly, it was shown that orienting to gaze occurs because of the perceived socia l re levance conveyed by eye direction (Chapter 3) and that once activated the social attention effect is resistant to cognitive top-down modulation (Chapter 2). Is eye direction the only central attentional cue that produces reflexive orienting? And if not, is there a distinction to be drawn between the effects elicited by different central c u e s ? In the three studies presented next in Sect ion II (Chapters 4-6) the attentional effects elicited by eye direction are compared to the attentional effects elicited by arrow cues . Chapter 4 reports an investigation that directly compares the orienting effect produced by nonpredict ive central gaze cues to the orienting effect produced by nonpredictive central arrow cues in three distinct populat ions: adult respondents, preschool chi ldren, and a split-brain patient. Surprisingly, eye direction and arrow direction produce behavioral ly indist inguishable effects in adults and preschool chi ldren. However, the attentional effects for the two cues dissociate for the split-brain patient, with only the face-process ing hemisphere orienting attention reflexively to gaze direction but both hemispheres orienting reflexively to arrow direction. Chapter 5 and Chapter 6 investigate the strength of the reflexive attention effect for gaze and arrow cues . The attentional literature indicates that reflexive attentional 46 effects are marked both by an ability to interrupt ongoing cognit ive activity (e.g., Muller & Rabbit, 1989) and by a resistance to modulation by changes in attentional set (e.g., F o l k e t a l , 1992). In Chapter 5, central eye gaze and arrow cues were counterpredict ive with regard to a target's likely location (e.g., a face looking to the left or an arrow pointing to the left predicted a target was likely to appear on the right). The results indicate that participants cannot help but attend initially toward where the eyes are looking. In contrast, any ev idence of reflexive orienting to arrow direction is abol ished. These data suggest that the attention effect triggered by eye direction is more strongly reflexive than the effect engaged by arrow direction. Chapter 6 examined whether the attentional effects of both central cues are influenced by changes in attentional set. A recent study by Pratt and Hommel (2003) found that attentional effects elicited by arrows cues are inf luenced by the attentional set created by a color congruency between the arrow cue and the target st imulus. Does a similar effect occur for eye g a z e ? In two experiments, participants were presented with spatially nonpredictive central arrow or gaze cues that either matched or mismatched a target's color (e.g., a central black arrow cue was fol lowed by either a black or a white target). The results indicate a clear dissociat ion between the orienting effects produced by arrow and gaze direction. Attentional orienting triggered by arrow direction is inf luenced by cue-target color congruency such that larger attentional effects emerge for congruent cue-target relations. In contrast, the orienting effect for gaze spans equal ly across congruent and incongruent cue-target color relations. Cons idered together the results of the three studies presented in Sect ion II indicate that central spatially nonpredict ive eye gaze cues are not unique in their ability to shift reflexive attention - central arrow cues also trigger reflexive orienting. However, 47 unlike the attention effects produced by central arrow cues the orienting elicited by eye direction appears to be more strongly reflexive such that it can co-occur with volitional orienting (Chapter 5) and its effects are resistant to changes in attentional set (Chapter 6). Thus, central eye gaze direction and arrow direction both can trigger reflexive orienting, although these two effects may be subserved by a different underlying neural architecture (Chapter 4) and orienting to gaze direction appears to be more strongly reflexive than orienting to arrows (Chapters 5 and 6). 48 References Driver, J . , Davis, G . , Ricciardel l i , P. , Kidd, P. , Maxwel l , E. , & Ba ron -Cohen , S . (1999). G a z e perception triggers visuospat ial orienting by adults in a reflexive manner. V isua l Cogni t ion, 6, 509-540. Fr iesen, C . K., & Kingstone, A . (1998). The eyes have it!: Ref lexive orienting is triggered by nonpredict ive gaze . Psychonomic Bulletin & Rev iew, 5, 490-495. Folk, C . L , Remington, R. W. & Johnston, J . C . (1992). Involuntary covert orienting is contingent on attentional control settings. Journal of Exper imental Psycho logy : Human Percept ion & Per formance, 18, 1030-1044. Hommel , B., Pratt, J . , Co lzato , L. & Godi jn, R. (2001). Symbol ic control of visual attention. Psycholog ica l Sc ience , 12, 360-365. Langton, S . R. H., & Bruce, V . (1999). Ref lexive socia l orienting. V isua l Cogni t ion, 6, 541-567. Langton, S , R, H. (2000). The mutual influence of head and gaze orientation in the analysis of socia l attention direction. Quarterly Journal of Exper imental Psycho logy, 53A, 825-845. Langton, S , R, H. & Bruce, V. (2000). Y o u must see the point: Automat ic processing of cues to the direction of socia l attention. Journal of Exper imental Psycho logy: Human Percept ion & Per formance, 26, 747-757. Muller, H. J . , & Rabbitt, P. M. A . (1989). Ref lexive and voluntary orienting of visual attention: time course of activation and resistance to interruption. Journal of Experimental Psycho logy : Human Percept ion and Per formance, 15, 315-330. Pratt, J . & Hommel , B. (2003) Symbo l ic control of v isual attention: The role of working memory and attentional control sett ings. Journal of Exper imental Psycho logy: Human Percept ion and Per formance, 29, 835-845 49 C H A P T E R 4 Are Eyes Spec ia l? It Depends on how you Look at it. A version of this chapter has been publ ished. Rist ic, J . , Fr iesen, C . K. & Kingstone, A . (2002). Are eyes spec ia l? It depends on how you look at it. Psychonomic Bulletin and Review, 9, 507-513. 50 In attentional cuing exper iments shorter reaction time (RT) to a target at a cued location versus a noncued location indicates that people have shifted their attention to the cued location. By varying the time interval between the cue and the target a temporal profile of the attentional effect can be establ ished (cf. Posner , 1980). Using modified vers ions of this standard paradigm, several recent studies have reported that spatially nonpredictive gaze direction facilitates R T to a target appear ing at the gazed-at location (Driver et al . , 1999; Fr iesen & Kingstone, 1998; Langton & Bruce, 1999). Specif ical ly, R T for a target at the cued (gazed-at) location is shorter than R T for a target at the noncued location 100 - 300 ms after onset of the gaze cue, and this effect d isappears by 1000 ms (Fr iesen & Kingstone, 1998). Fr iesen and Kingstone (1998) proposed that this reflexive orienting to gaze direction may represent an attentional process that is unique to biologically relevant stimuli, and that as such , it may be subserved by brain regions specif ic to the processing of faces and eyes . In support of this hypothesis, Kingstone, Fr iesen and G a z z a n i g a (2000) found that when individuals who have had their cerebral hemispheres surgically d isconnected (split-brain patients) were presented with nonpredict ive gaze stimuli, only the hemisphere specia l ized for face processing directed attention reflexively to the gazed-at location. Remarkably , however, there is no publ ished report directly compar ing attentional orienting to spatially nonpredict ive gaze cues (biologically relevant) with attentional orienting to spatially nonpredict ive symbol ic cues , such as arrows (biologically irrelevant). S u c h a compar ison represents a crucial test of the "eyes are spec ia l " posit ion. It is possible that this test has not been performed because the convent ional w isdom in the field is that spatially nonpredict ive central arrow cues will not produce reflexive orienting (cf. Langton et al . , 2000, p. 55). This view s tems largely from a 51 c lass ic study by Jon ides (1981, Experiment 2) that required subjects to search a briefly presented array of letters for the target L or R. Before the array appeared, a central arrow cue was f lashed momentari ly at fixation. The arrow pointed randomly at one of the letter locations. Resul ts indicated that if subjects were told to ignore the arrow then orienting to the cued location was absent. This suggests that a nonpredict ive arrow cue does not trigger reflexive attention. However, we will show that this finding does not hold when the task is target detection and the arrow cue remains present (seeTipples, in press for a similar result). Thus , it is an open question as to how reflexive orienting to gaze and arrow cues compare. W e addressed this issue in three ways . First, we tested the s a m e adult observers with nonpredict ive gaze and arrow cues across a range of cue-target intervals to compare the strength, and the temporal profile, of orienting to biologically relevant and irrelevant directional cues . S e c o n d , we tested 4- and 5-year-old children with these s a m e condit ions. G iven that infants are predisposed to attend to faces and eyes (e.g., Maurer, 1985) and begin to follow gaze direction within the first year (e.g., D'Entremont, Hains, & Muir, 1997), we expected that nonpredict ive gaze would produce reflexive orienting in young chi ldren. In contrast, given chi ldren's more limited exper ience with arrow stimuli, we expected that nonpredictive arrow cues would produce smal ler orienting effects or none at all. Finally, we tested split-brain patient J . W . with nonpredictive arrow cues . Wou ld the lateralization found for nonpredict ive gaze cues (Kingstone et a l . , 2000) a lso occur for nonpredictive arrow c u e s ? If not, the implication is that the cortical brain mechan isms subserv ing reflexive orienting to biologically irrelevant stimuli are distinct from those subserv ing reflexive orienting to gaze stimuli. 52 Experiment 1 The Fr iesen and Kingstone (1998) paradigm was modified to compare the attentional effects of nonpredictive gaze and arrow cues . B e c a u s e we planned to apply precisely the same paradigm to preschool chi ldren in Exper iment 2, the targets were pictures of a snowman and a cat. Method Part icipants Nineteen psychology undergraduate students from the University of British Co lumbia participated for course credit. Apparatus A 3200c Macintosh Power Book presented stimuli on a 12-inch black and white monitor. Part icipants were seated approximately 57 cm from the monitor. Target detection R T was measured as the time interval between target onset and pressing the spacebar (marked with red tape). Stimuli Stimuli and trial sequences are illustrated in Figure 4.1. Al l stimuli were black drawings on a white background. For gaze cues , the central fixation st imulus was a line drawing of a happy face subtending 6° . The face contained two 0.8° circles representing eyes , a 0.2° circle centered within the face outline representing a nose, and a curved 2.6° long line representing a smil ing mouth. Black filled-in circles representing pupils appeared in the eyes . The pupils were centered vertically in the eyes , and just touched the left or right eye outline. The pupils measured 0.5°, and the distance between the eyes was 1° when measured from the center of each eye. For arrow cues , the fixation st imulus was a horizontal line centered on the sc reen , 1.9° in length. A n arrow head and an arrow tail appeared at the ends of the central line, both 53 pointing left or both pointing right. E a c h of the two lines compris ing an arrow head or tail measured 0.5°, and the length of an arrow, from the tip of the arrow head to the ends of the tail, was 2.5°. The two target stimuli were drawings of a snowman and cat. The cat was 2.5° wide x 3° high; the snowman was 2.5° x 4° . Targets appeared 5° to the left or right as measured from the center of the face or arrow to the center of the target. Des ign A trial began by presenting a face with blank eyes or a straight line for 936 ms. Then pupils or an arrow appeared . A target appeared on the left or right s ide of the screen 195, 600, or 1005 ms after cue onset. Stimuli remained on the sc reen until a response was made or 3800 ms had e lapsed, whichever came first. The intertrial interval was 808 ms. On cued trials the target appeared at the location towards which the cue was directed, and on uncued trials the target appeared at the other location. Part icipants completed four blocks of 42 trials, two consecut ive blocks with gaze cues and two consecut ive blocks with arrow cues . C u e order was counterbalanced across participants (10 received gaze cues first). C u e direction, target location, target identity, and cue-target stimulus onset asynchrony ( S O A ) were selected randomly and equally. Four catch trials, in which no target was presented, occurred randomly in each block. Procedure Part icipants first received a descript ion of the trial sequence and completed several practice trials. They were told that gaze and arrow direction did not predict target location or identity. Finally, participants were instructed to press the spacebar quickly and accurately when the target appeared, and to maintain central fixated during each block. 54 Resul ts Med ian R T s were calculated for each participant. The interparticipant means of these median R T s are illustrated in Figure 4. 2. Figure 4.2 shows that for both gaze and arrow cues R T was shorter at the cued than at the uncued location at the 195 ms S O A by approximately the s a m e amount (21 ms and 22 ms, respectively). A s S O A lengthened the R T difference between cued and uncued locations decreased and R T became shorter overall (with a slight R T increase at the 1005 ms S O A — a c lass ic cue-target foreperiod effect (Bertelson, 1967; Mowrer, 1940). These observat ions were conf irmed by a three-way analys is of var iance ( A N O V A ) with cue type (gaze, arrow), cue validity (cued, uncued), and S O A (195 ms, 600 ms, 1005 ms) as within-subject factors. There were significant main effects for validity [F (1, 18) = 39.54 p < .0001] and S O A [F (2, 18) = 40.58 p < .0001], reflecting that R T was shorter at cued vs . uncued locations, and shorter overall at the longer S O A s . There was also a significant validity and S O A interaction [F (2, 18) = 5.19, p < .02], representing that the difference between cued and uncued locations decreased as S O A lengthened. P lanned contrasts revealed that the cuing effect was signficant at all S O A s [all F s > 5, all ps <.05]. Finally, the interaction between cue type and S O A was significant [F (2, 18) = 13.93, p < .0001], reflecting that R T was initially longer for an arrow cue than a gaze cue. No other effects approached signi f icance. In particular, there was no significant main effect for cue type [F(1,18) = 1.36, p>.25], or interaction involving cue type and validity (cue x validity, [F (1,18) = 1.79, p = .20]; cue x S O A x validity, [F < 1]). There were no incorrect key presses on target-present trials. Fa lse a larms were classif ied as errors and were exc luded from the analysis. The false alarm rate on catch 55 trials was 2.4% in the gaze condit ion and 3 .2% in the arrow condit ion. There was no significant effect of cue type [F <1]. D iscuss ion There were two key findings in Exper iment 1. First, the basic pattern reported by Fr iesen and Kingstone (1998) and others (Driver et a l . , 1999; Langton & Bruce, 1999) was repl icated. That is, nonpredictive gaze direction triggered a rapid shift of attention to the gazed-at location. The result was R T facilitation at the gazed-at location 195 ms after onset of the social cue, with the facilitatory effect decl ining as the cue-target S O A approached 1000 ms. S e c o n d , this effect was a lso observed for nonpredictive arrow cues . In other words, the reflexive attentional effect is not unique to biologically relevant gaze cues . O n this point the data are absolutely unequivocal . Nonpredict ive arrows trigger a reflexive shift of attention to the cued location in a manner that is effectively indist inguishable from gazed-tr iggered orienting. Indeed, the only difference between the two cues was that initially R T was longer for arrows than eyes , reflecting perhaps the fact that gaze is more alerting than an arrow. Finally, note that Tipples (in press) reported reflexive orienting to peripheral nonpredictive arrow cues . However, as he noted, this effect might be an artifact of the arrow cues being presented peripherally, and/or bilaterally. The present data rule out these possibil i t ies and as such , to our knowledge, they represent the first c lear demonstrat ion that a nonpredictive central arrow cue will trigger reflexive orienting to a cued peripheral location. Experiment 2 In Experiment 2 we tested 3- 5-year-old preschool children with exactly the same stimulus condit ions as the ones that were appl ied to the adults in Exper iment 1. 56 A s noted in the introduction of this paper there is a t remendous amount of ev idence indicating that infants are pred isposed to preferentially process faces and eyes , and that within their first year they direct their attention to where others are looking (e.g., D'Entremont, Hains, & Muir, 1997; Maurer, 1985). This suggested to us that nonpredictive gaze cues would have greater attentional sa l iency for children than biologically irrelevant symbol ic stimuli such as arrows, with which children have less exper ience. Thus we predicted that nonpredictive gaze cues would produce orienting effects in preschool chi ldren, and that nonpredictive arrow cues would produce either smaller orienting effects than nonpredictive gaze cues , or no orienting effects at all. It is a lso worth noting that a compar ison of the adults' and children's results would provide a novel way to test the extent to which attentional orienting to nonpredictive gaze and arrow cues is reflexive in nature. Specif ical ly, there have been recent suggest ions that reflexive orienting to gaze cues in adults may be a learned volitional process (Vecera, 2000, unpubl ished ms). B e c a u s e children younger than 8 years demonstrate adult-like reflexive attention effects, but muted volitional attention effects in peripheral cuing exper iments (Brodeur, Trick, & Enns , 1997), we reasoned that if orienting to biologically relevant (or irrelevant) stimuli was a volitional process, then the orienting effects should be smal ler for the preschoolers than for the adults. Method Part icipants Twenty-eight preschool chi ldren were recruited from two Vancouve r daycare facilit ies, and parental permission was obtained for each chi ld. Nine children failed to complete the experiment. A g e s ranged from 3 years, 9 months to 5 years, 10 months (mean age 4 years, 8 months). Apparatus, des ign, and procedure 57 A s in Exper iment 1, with two except ions: (1) each cue condit ion was composed of one block of 42 trials, and (2) two experimenters were present when the experiment was conducted. One experimenter ensured that central fixation was maintained and the other that the correct response key w a s pressed. Extra care was taken in explaining that the direction of the eyes and arrows did not predict where the target would appear or what target would appear. Resul ts Figure 4.2 shows that for both gaze and arrow cues R T was shorter at the cued than at the uncued location at the 195 ms S O A , and that this effect persisted across all S O A s (although disappear ing temporari ly at the 600 ms S O A for gaze cues) . A s S O A lengthened R T became shorter overall (the cue-target foreperiod effect). Mirroring the adult data in Exper iment 1, a three-way A N O V A revealed significant main effects for cue validity [F (1, 18) = 18.88 p < .0005] and S O A [F (2, 18) = 4.24 p < .03]. There was no significant main effect for cue type [F < .01], and no significant interactions [all F s < 1.2, all ps > .31]. Incorrect key presses on target-present trials were 0.3 % for the gaze condit ion and 0 .8% for the arrow condit ion, and these did not vary as a function of validity or S O A [all Fs<1]. The false alarm rate on catch trials was 4 1 . 5 % for the gaze condit ion and 4 0 % for the arrow condition (with a nonsignif icant difference between cue condit ions, [F<1]). These rates are obviously very high; however, it is important to note that they do not compromise the effects observed on target present trials. That is because false alarm responses normally occurred long after the gaze and arrow cues were presented (mean R T of 1841 ms and 1731 ms, respectively), with these long R T s falling well outside the latencies observed on target present trials. Thus the high false alarm rates 58 merely reflect the fact that the children had difficulty inhibiting a key press response for the full duration of a catch trial, i.e., they were not target anticipations. Finally, we performed a four-way A N O V A to compare the adults' and chi ldren's R T data. C u e type, cue validity, and S O A were included as within-subject var iables, and age group (adults vs . children) was included as a between-subjects variable. Consistent with previous f indings (e.g., Enns & Akhtar, 1989), adults' R T was shorter overall [F (1, 36) = 192.68, p < .0001]. The main effect for cue type was not significant [F (1, 36) = .050, p >.82], and the main effects for S O A [F (2, 36) = 11.50, p < .0001] and validity [F (1. 36) = 31.44, p < .0001] were highly significant. The validity x age group interaction was also significant [F (1, 36) = 8.3, p < .007], indicating that the chi ldren showed a larger cuing effect than adults. Of course whether children would show a larger cuing effect if they were not significantly s lower than adults must still be determined. No other interaction was significant [all F s < 1.6, all ps > .20]. D iscuss ion The results from Exper iment 2 are clear-cut. Consistent with our expected finding, children oriented attention reflexively to the location cued by a nonpredictive gaze st imulus. Unexpectedly, however, children a lso oriented attention reflexively to the location cued by a nonpredict ive arrow stimulus, and this effect was statistically indist inguishable from the gaze effect. T h e s e data replicate for chi ldren what we observed for adults in Exper iment 1, and again bring into question the notion that the effects of a nonpredictive gaze st imulus are unique to biologically relevant stimuli. Experiment 3 The results of the first two exper iments strongly indicate that reflexive orienting to a biologically relevant nonpredict ive gaze cue has a behavioral effect that is 59 indist inguishable from reflexive orienting to a biologically irrelevant nonpredict ive arrow cue. Does this mean that the two types of cue activate the same brain pathways? A recent study by Kingstone et al . (2000) revealed that split-brain patients shift their attention reflexively to a gazed-at location only when the gaze cue projects to the hemisphere that is specia l ized for the processing of face stimuli. This suggests that reflexive attention to gaze direction reflects an interaction between neurons in the temporal cortex of the hemisphere specia l ized for process ing faces and eyes , and neurons in the parietal cortex responsible for orienting spatial attention (Harries & Perrett, 1991; Hoffman & Haxby, 2000; P u c e et al . 1998; Wicker et a l . , 1998 ). Would a split-brain patient show a similar lateralization of function for nonpredictive ar rows? If the neural mechan isms that are responsible for orienting attention to nonpredictive arrow cues are the same as the neural mechan isms responsible for orienting attention reflexively to nonpredictive gaze cues , then J . W . should demonstrate reflexive orienting in the right hemisphere but not the left hemisphere. Method Participant Cal losotomy patient J . W . is a 46-year-old male who suffered from intractable epi lepsy beginning in 1972. Both hemispheres comprehend language, although verbal and written language output is lateralized to the left hemisphere. This patient has participated in numerous behavioral investigations and is well known for holding central fixation on instruction. S e e G a z z a n i g a , N a s s , Reeves , and Roberts (1984) for a detailed descript ion of this patient. Stimuli 60 Stimuli were controlled by an Apple Macintosh PowerBook 180c computer connected to a 14-inch monitor. The stimuli, illustrated in Figure 4.3, were black and the background was white. The arrow stimuli were the s a m e as in Exper iments 1 and 2, and were posit ioned 2.4° to the left and right of fixation. The target was an asterisk that subtended 0.7°, and a lways appeared 4.2° away from the central fixation cross (which subtended 0.3°). Procedure J . W . was centered with respect to the screen and keyboard, and central fixation was held without difficulty throughout each block of trials. Twenty (20) practice trials preceded two sets of 10 blocks of 64 test trials, for a total of 1280 test trials. J . W . was informed repeatedly, and understood, that arrow direction did not predict where the target would appear. He was strongly encouraged to respond as quickly and as accurately as he could, by pressing a left-hand key ("Z") when the target was presented to the left visual field (LVF ; right hemisphere), and a right-hand key ("/") when the target was presented to the right visual field ( R V F ; left hemisphere). Figure 4.3 presents an example sequence of events for a trial. Two vertical l ines were presented concurrently to the left and right of fixation. 675 ms later arrow heads and tails appeared , creating arrows that pointed up or down. After 105 ms or 600 ms, a target appeared above or below one of the arrows. Arrow direction, target location, and cue-target S O A were selected randomly and equal ly within each block. Stimuli remained on the screen until a response was made or 2700 ms had e lapsed , whichever came first. The intertrial interval was 675 ms. Note that this procedure dupl icated Kingstone et al . (2000) except that arrow cues replaced gaze cues. 61 Resul ts Figure 4.3 shows that for both L V F (right hemisphere) and R V F (left hemisphere) targets, R T was shorter at the cued than the uncued location at 105 and 600 ms S O A s . Addit ionally, as S O A increased R T became shorter overal l (the cue-target foreperiod effect), with this effect being greater for L V F than R V F targets. These arrow cue data contrast dramatically with the Kingstone et al . (2000) finding that gaze direction triggered reflexive orienting only for L V F (right hemisphere) targets. R T and accuracy data were subjected to an A N O V A with cue validity, S O A , and target field as factors. R T analys is revealed that all main effects were significant [all F s > 8.93, all ps < .01]. There was also a significant interaction between S O A and target field [F (1, 1251) = 5.38, p <02] reflecting the greater foreperiod effect for L V F than R V F targets. Importantly, there were no other significant interactions [all F s <1.5, all ps > 20]. Error analysis produced no significant effects [all Fs < 1.2, all ps >.35]. D iscuss ion The finding that nonpredictive arrow direction produces a cuing effect in both hemispheres of split-brain patient J .W. , at both S O A s , contrasts with Kingstone et al . (2000) finding that nonpredictive gaze direction produces a rapid, and short-l ived attention effect that is lateralized to J .W. ' s face/gaze processing right hemisphere. Together these two findings strongly suggest that the neural mechan isms that subserve a reflexive shift of attention in response to nonpredict ive gaze direction are fundamental ly different from the mechan isms that subserve reflexive orienting in response to nonpredictive arrows. This agrees with current work indicating that there exists a distinct brain region that is specia l ized for process ing biologically relevant directional face and gaze information, which is not activated by inanimate biologically 62 irrelevant directional information, such as arrows (see Kanwisher , Downing, Epste in, & Kourtzi , 2001 for a review). Genera l D iscuss ion Three exper iments were conducted to examine whether attentional orienting triggered by spatially nonpredict ive and biologically relevant gaze cues differs from attentional orienting triggered by spatially nonpredict ive and biologically irrelevant cues , such as arrows. The results were unambiguous. Our first two experiments found that the behavioral effects of nonpredict ive gaze cues and arrow cues were significant and equivalent across adults and chi ldren. At first pass these data would appear to compromise the "eyes are spec ia l " position put forward by Fr iesen and Kingstone (1998), Langton and Bruce (1999) and Driver et a l . (1999). That is, the position that the reflexive attentional orienting observed for gaze cues reflects an attentional network that is qualitatively distinct from attentional orienting triggered by biologically irrelevant stimuli. In support of this position, Kingstone et al . (2000) found that when nonpredict ive gaze cues are presented to split-brain patient J . W. , only the right ( face/gaze processing) hemisphere attends reflexively to the gazed-at location. Whi le it is tempting to conclude that this effect is specif ic to the processing of biologically relevant face stimuli, an alternative possibil ity is that any directional st imulus, biologically relevant or not, will trigger a lateralized reflexive shift of attention (see Hommel l , Pratt, Co lzato & Godi jn, 2001 for a similar considerat ion with regard to nonpredictive arrows and words). Indeed, the behavioral equivalence of gaze and arrows observed in Exper iments 1 and 2 supports this alternative. It is therefore new and significant to d iscover in Exper iment 3 that nonpredictive arrows produce reflexive orienting in both hemispheres of split-brain 63 patient J .W. , in contrast to the lateralized effect obtained with nonpredict ive gaze . In this very important sense biologically relevant gaze cues are spec ia l . Note that the split-brain data do not indicate simply that any index of attentional orienting to gaze direction will be lateral ized to the hemisphere that is preferentially b iased to processing face and gaze information. Rather, the key is whether the attentional orienting is reflexive (triggered by a nonpredictive gaze cue) or volitional (triggered by a predictive cue; see Danziger & Kingstone, 2000, for a recent review of exogenous vs . endogenous orienting). To demonstrate this point Kingstone et al . (2000) conducted a control study wherein they repeated the procedure used in the present Exper iment 3 but with gaze stimuli that predicted where the target stimulus was likely to appear, i.e., in either visual field the target appeared at a gazed-at location on 7 5 % of the trials and at a non-gazed-at location on 2 5 % of the trials. Here both hemispheres attended volitionally to the predicted gazed-at location. Thus the lateralization of reflexive attention to nonpredictive gaze direction reflects an interaction between gaze processing and the predictive (attentional) value of the gaze cue. In sum, our results with J . W . (present study; Kingstone et al . , 2000) indicate that reflexive orienting to nonpredict ive gaze is subserved by a neural sys tem that is qualitatively unique both from the system that supports reflexive attention to biologically irrelevant stimuli and from the system that supports volitional orienting to predictive gaze direction. In this important way, reflexive orienting to eye direction is spec ia l . Moreover, a recent study with healthy adult observers indicates that reflexive and volitional orienting to gaze direction are behavioural ly separable. Fr iesen, Rist ic, and Kingstone (submitted) d iscovered that if eye direction predicts that a target will appear at a nonqazed-at location, attention is committed reflexively to the gazed-at location concurrent with volitonal orienting to the predicted location. Counterpredict ive arrows on 64 the other hand appear to produce only reflexive or volitional effects. Thus we find that even in healthy observers behavioural dif ferences between eyes and arrows may emerge, consistent with the finding of the present study that eyes are indeed specia l . 65 Figure 4 .1 . Figure 4.1. Illustration (not to scale) of stimuli and sample sequence of events for Experiment 1 and Experiment 2. For both experiments, the start of each trial was signaled by the presentation of either a face with blank eyes or a straight line. 936 ms later the pupils (looking left or right) or an arrow (pointing left or right) appeared. A target (snowman or cat) was presented on the left or right side of the fixation stimulus, 195, 600, or 1005 ms after cue onset. Both the central cue and the target remained on the screen until a response was made or 3800 ms had elapsed, whichever came first. Response time (RT) was measured from the onset of the target. 66 Figure 4.2. Experiment!: Adults Experiment 2: Children 440 420) -o400i o 0 38CH CO f 360 [£340-300--0-195 ms S O A -D-600 m s S O A - A - 1005 ms S O A G A Z E A A n • 740 720 70O 680 660 640 620 600--0-195 m s S O A -•-600 m s S O A -A-1005 ms S O A 0.9 % G A Z g 0.9% 440, A R R O W 740, A R R O W _420 co 720 70a 0.7% lliseci CO oo 68a JL360 -° 66a 1.7 % ^340 o- " 64a 32a A 62a 1.7 a / a t ^ \ ^ ^ ^ 30O 6oa 440, GAZE+ARROW 740, 42a 720 co •g 4oa o 0 38a CO 70O 680 1 36a 660 or 34a ^ 640 32a A — — — — * 620 30a •? : ° 60O Cued Uncued C u e Cued Uncued C u e Figure 4.2. Mean response time (RT) in milliseconds (ms) for Experiment 1 (adult participants) and Experiment 2 (child participants), as a function of cue-target stimulus onset asynchrony (SOA) and cue validity. The top row shows performance for gaze cues, the middle row shows performance for arrow cues, and the bottom row shows performance collapsed across gaze and arrow cue conditions. Error rates (%) that are not zero are shown. 67 Figure 4.3. Left Visual Field (Right Hemisphere) 4201 410 400 390 380 370 360 350 340 105 ms S O A -^CT625% 1.25% - 600 ms S O A 1.25% i 1 Cued Uncued Cue 2700 ms 1 - 1 105 or 600 ms 1 • 1 675 ms START Right Visual Field (Left Hemisphere) 4201 410 _ 400 CO 1 390 o .«? 380 or 370 360 350 340 . 105 ms S O A ^ ^ 0 % . 3.125% " 600 ms S O A / 0.625% 1 Cued Uncued Cue Figure 4.3. Illustration (not to scale) of stimuli, sample sequence of events, and mean response time (RT) for Experiment 3 (split-brain participant J.W.), as a function of cue-target SOA, cue validity, and visual field. The error rates (%) for all conditions are shown. See text for procedural details. 68 References Berte lson, P. (1967). The time course of preparation. Quarterly Journal of Experimental Psycho logy. 19, 272-279. Brodeur, A . , D., Trick, M. , I, & Enns , J . T. (1997). Select ive attention over the l i fespan. In Burack A . J . & Enns J . T. (Eds.) . Attention, development, and psychopathology (pp. 74-97). New York: Guilford P ress . D'Entremont, B., Hains, S . M. J . & Muir, D. W. (1997). A demonstrat ion of gaze following in 3-to-6 month-olds. Infant behavior and development, 20, 569-572. Danziger, S . & Kingstone, A . (1999). Unmask ing the inhibition of return phenomenon. Percept ion & Psvchophys ics . 61 , 1024-1037. Driver, J . , Davis, G . , Ricciardel l i , P. , Kidd, P. , Maxwel l , E., & Baron -Cohen , S . (1999). G a z e perception triggers visuospat ial orienting by adults in a reflexive manner. V isual Cogni t ion. 6 , 509-540. Enns , J . T., & Akhtar, N. (1989). Relat ions between covert orienting and filtering in the development of visual attention. Journal of Exper imental Chi ld Psycho logy, 48, 315-334. Fr iesen, C . K., & Kingstone, A . (1998). The eyes have it!: Ref lexive orienting is triggered by nonpredictive gaze . Psvchonomic Bulletin and Rev iew, 5, 490-495. Fr iesen C . K., Rist ic, J . & Kingstone A . (submitted). Ref lexive and volitional orienting to directional cues : Separab le attention effects unique to biologically relevant gaze stimuli. Journal of Exper imental Psycho logy: Human Percept ion and Per formance. G a z z a n i g a , M. S . , N a s s , R., Reeves , A . , & Roberts, D. (1984). Neurologic perspect ives on right hemisphere language following surgical sect ion of corpus ca l losum. Seminars in Neurology, 13, 536-540. Harries, M. , & Perret, D. I. (1991). V isua l process ing of faces in temporal cortex: Physio logical ev idence for a modular organizat ion and possible anatomical correlates. Journal of Cognit ive Neurosc ience, 3, 9-24. Hoffman, E. A . & Haxby, J . V . (2000). Distinct representat ions of eye gaze and identity in the distributed human neural sys tem for face perception. Nature Neurosc ience, 3, 80 - 84. Hommel , B., Pratt, J . , Co lzato , L. & Godi jn, R. (2001). Symbol ic control of visual attention. Psycholog ica l Sc ience , 12, 360-365. Jon ides , J . (1981). Voluntary vs . automatic control over the mind's eye 's movement. In J . B. Long & A . D. Baddley (Eds.) , Attention and performance. IX (pp. 187-203). Hi l lsdale, N J : Er lbaum. Kanwisher , N., Downing, P. , Epste in, R., &Kourtzi , Z . (2001). Funct ional neuroimaging of v isual recognit ion. In R. C a b e z a & A . Kingstone (Eds.), Handbook of functional neuroimaging of cognit ion, (pp. 109-151). Cambr idge, M A : MIT Press . Kingstone, A . , Fr iesen, C . K., & G a z z a n i g a , M. S . (2000). Ref lexive joint attention depends on lateralized cortical connect ions. Psycholog ica l Sc ience , 11, 159-165. Langton, S . R. H., & Bruce, V . (1999). Reflexive socia l orienting. V isua l Cogni t ion, 6, 541-567. Langton, S . R. H., Watt, R. J . , & Bruce, V . (2000). Do the eyes have it? C u e s to the direction of socia l attention. Trends in Cognit ive Sc iences , 4, 50-59. Maurer, D. (1985).Infants' perception of facedness . In T. M. Field & N. A . Fox (Eds.). Soc ia l perception in infants (pp. 73-100). Norwood, N J : Ab lex . Mowrer, O . H. (1940). Preparatory set (Expectancy) - S o m e methods of measurements . Psycholog ica l review monograph, 52 (Whole No. 233). Posner , M. I. (1980). Orienting of attention. Quarterly Journal of Exper imental Psycho logy, 32, 3-25. P u c e , A . , Al l ison, T., Bent in, S . , Gore , J . C , & McCar thy , G . (1998). Tempora l cortex activation in humans viewing eye and mouth movements . The Journal of Neurosc ience, 18, 2188-2199. Tipples, J . (in press). Eye gaze is not unique: Automat ic orienting in response to noninformative arrows. Psychonomic Bulletin and Review. V e c e r a , S . (2000). The eyes may not have it: Interrupting the al location of "social attention". Unpubl ished manuscript. Wicker , B. F., Miche l , F., Henaff, M. , & Decety, J . (1998). Brain regions involved in the perception of gaze : A P E T study. Neuro image, 8, 221-227. 71 C H A P T E R 5 Attentional Effects of Counterpredict ive G a z e and Arrow C u e s A version of this chapter has been publ ished. Fr iesen, C . K., Rist ic, J . & Kingstone, A . (2004). Attentional effects of counterpredictive gaze and arrow cues . Journal of Experimental Psychology: Human Perception and Performance, 30, 319-329. 72 Behavioral studies with healthy adults have indicated that the tendency to move attention to where someone else is looking is so fundamental that people will attend automatically to a location gazed at by a face on a computer sc reen, even when gaze direction does not predict where a target item may appear (e.g., Driver et al . , 1999; Fr iesen & Kingstone, 1998; Langton & Bruce, 1999). More recently, a study by Kingstone, Fr iesen and G a z z a n i g a (2000) revealed that this effect is lateralized to the hemisphere specia l ized for processing face and gaze information. These findings, coupled with the observat ion that gaze direction can convey a broad range of important social s ignals, have led to the suggest ion that orienting to gaze direction may represent a specia l form of attention (Langton, Watt, & Bruce, 2000; Kingstone, Smi lek, Rist ic, Fr iesen & Eas twood, 2003). The present study investigated this hypothesis by examining attentional orienting in response to directional stimuli that were either gaze cues or arrow cues . In their original gaze study with adults, Fr iesen and Kingstone (1998) reported that when a schemat ic face was presented in the center of a computer sc reen and the gaze direction of the face was known to be spatially nonpredict ive, adults were nevertheless faster to detect, local ize, and identify a target st imulus if it appeared at the location that the face was looking at rather than at a nongazed-at location. A n equally interesting result was that this facilitatory effect of gaze direction emerged soon after the schemat ic eyes were presented — at a cue-target stimulus onset asynchrony ( S O A ) of 105 mi l l iseconds (ms) — and then persisted across S O A s of 300 and 600 ms before d isappear ing by a 1005 ms S O A . Both the rapid onset of the facilitation effect, and the fact that it occurred in response to a nonpredict ive st imulus, are hal lmarks of reflexive attentional orienting (Cheal & Lyon, 1991; Mu l le r& Rabbitt, 1989; Jon ides , 1981). This 73 suggested to Fr iesen and Kingstone that they were measur ing a reflexive attentional phenomenon. However, they also noted that orienting to gaze direction did not exhibit all the characterist ics normally assoc ia ted with reflexive shifts of attention. For instance, in their study the attentional shift to a peripheral location was triggered by a spatially nonpredictive st imulus (the eyes) presented at central fixation. In contrast, reflexive orienting is normally produced by presenting a spatially nonpredictive transient event, such as the brightening of a box, at a peripheral location where a target might appear (e.g., Posner & C o h e n , 1984; Posner , C o h e n , & Rafa l , 1982). Another difference was that Fr iesen and Kingstone found that orienting to gaze direction persisted well beyond a cue-target S O A of 500 ms. The reflexive orienting effect produced by nonpredict ive peripheral cues d isappears when the cue-target S O A exceeds approximately 300 ms (Klein, Kingstone & Pontefract, 1998). A final difference was that when the facilitory effect of gaze direction d isappeared, it was never replaced by the inhibition of return (IOR) effect, i.e., an increase in response time (RT) for targets appear ing at the cued location. This contrasts with spatially nonpredictive peripheral cues , in which the short-lived early facilitation effect at the cued location is typically replaced by an IOR effect at longer S O A s (Posner & C o h e n , 1984; Posner , Rafa l , Choate , & Vaughan , 1985; for a review, see Klein 2000). Cons idered together, these dif ferences suggested to Fr iesen and Kingstone that attention to gaze direction might represent a new, and different, type of reflexive orienting. Similar f indings, and conclus ions, were put forward by Langton and Bruce (1999) and Driver et al . (1999). In addition, each of these two studies examined volitional orienting to gaze direction by testing performance when gaze direction predicted where a target st imulus w a s likely to appear. Langton and Bruce (1999; Exper iment 3) 74 examined volitional orienting by presenting an image of a real face in the center of the computer sc reen . The face could be turned either to the left, to the right, up, or down. Subjects were informed that the target stimulus would appear 7 5 % of the time at the location that the head and eyes were directed toward (the cued location), and 2 5 % of the time at one of the other three uncued target locations. Resul ts indicated that R T was facilitated for targets appear ing at the cued location both when the cue-target S O A was short (100 ms) and when it was long (1000 ms). Langton and Bruce suggested that the facilitation observed at the short S O A reflected a reflexive shift of attention to the g a z e d -at location (because this effect was a lso observed at the short S O A in their first two exper iments with nonpredict ive gaze) , and that the facilitation effect observed at the long S O A reflected voluntary orienting to the gazed-at location (because the attentional effect had d isappeared at this long S O A in their nonpredictive gaze experiments). This account is both reasonable and consistent with the data. However, because Langton and Bruce only sampled performance at two temporal extremes — a short 100 ms S O A at which gaze-tr iggered reflexive orienting is often observed, and a longer 1000 ms S O A at which gaze-tr iggered reflexive orienting is often absent — their results do not indicate when voluntary orienting in response to the predictive cue emerged, or more specif ical ly, whether this voluntary orienting effect replaces reflexive orienting. Driver et a l . (1999, Exper iment 3) tested the reflexivity of orienting to gaze direction by making the gaze cue counterpredict ive with respect to where a target was likely to appear. Observers were presented with an image of a real face pointed straight ahead but with eyes gazing to the left or right. They were informed that when the eyes looked to the left, the target would appear on the right 8 0 % of the time, and vice versa . R T performance was sampled at 100, 300, and 700 ms cue-target S O A s . No effects of gaze direction were observed at the shortest S O A of 100 ms; however, at the 300 ms 75 S O A , R T was shorter at the location that the eyes were directed toward (where the target was unlikely to appear), and at the 700 ms S O A there was a nonsignif icant trend for R T to be shorter at the location opposite to where the eyes were directed (where the target was likely to appear). In keeping with Langton and Bruce (1999), Driver et al . suggested that at the shorter 300 ms S O A , attention was reflexively committed to where the eyes were looking; whereas at the longer 700 ms S O A , attention was voluntarily shifted to the location where the target was likely to appear (in this case , the opposite location). Al though this is a plausible interpretation of the data, the fact remains that performance was never significantly faster at the nongazed-at (but likely) target location, and therefore the ev idence does not provide strong support for the view that the reflexive orienting observed at 300 ms S O A was replaced by volitional orienting at 700 ms S O A . One alternative interpretation is that with counterpredictive gaze cues , the conflict between the attentional effects of gaze direction and the task requirement to shift attention in the opposite direction somehow diminishes or delays volitional orienting (which usual ly emerges at cue-target S O A s of about 500 ms or less (Danziger & Kingstone, 1999; Muller & Rabbitt, 1989)). A second alternative is that reflexive orienting to gaze direction and voluntary orienting to the likely target location were both occurring at the 700 ms S O A . If this were the case , the overall result might be to facilitate R T performance both at the gazed-at location (because of reflexive orienting) and at the likely location (because of volitional orienting), thereby reducing or eliminating any significant dif ferences between these two locations. Note that this alternative interpretation is reasonable because both Driver et al . (1999, Exper iments 1 and 2) and Fr iesen and Kingstone (1998) demonstrated originally with nonpredict ive gaze cues that reflexive orienting can be observed with cue-target S O A s as long as 600 - 700 ms. 76 Consider ing the data from these studies as a whole, it becomes clear that they do not provide a clear picture of reflexive and voluntary orienting in response to central gaze direction cues . The results of Langton and Bruce (1999) indicate that participants can orient attention both reflexively and volitionally in response to predictive face/gaze directional cues . But they do not reveal whether volitional orienting rep laces, or over laps with, reflexive orienting. Similarly, the findings of Driver et a l . (1999) with counterpredictive gaze cues indicate that at a short S O A of 300 ms, attention is oriented reflexively to the gazed-at location even when participants have an incentive to shift their attention in the direction opposite to where the eyes are looking, suggest ing that orienting to gaze direction may be strongly reflexive. And the trend towards a response time advantage for targets appear ing at the predicted location at the longer 700 ms S O A suggests that some volitional orienting might have been occurr ing at this longer S O A . However, for the reasons just d i scussed , it is equivocal whether the nonsignificant benefit for predicted (but not gazed-at) targets at 700 ms S O A simply reflects weak or de layed volitional orienting, or whether reflexive orienting and volitional orienting were co-occurr ing at this longer S O A . In Exper iment 1 of the present study we used counterpredict ive gaze cues in an improved design that al lowed us to isolate reflexive orienting to a gazed-at location from voluntary orienting to a predicted location, and to examine the t imecourse of orienting to counterpredictive gaze direction ac ross a wide range of S O A s . In Exper iment 2 we examined the attentional effects of counterpredictive arrows, in order to investigate whether the effects obtained with gaze cues in Exper iment 1 are a lso observed with another common directional cue, i.e., an arrow. Exper iment 1 77 In the Driver et a l . counterpredictive gaze experiment (1999, Exper iment 3), participants oriented reflexively to a gazed-at but unlikely target location at a short S O A of 300 ms, but there was no clear ev idence of a switch to volitional orienting to a likely target location at a longer S O A of 700 ms. Two possible reasons for the absence of significant volitional orienting were advanced above. One possibility is that when gaze direction is counterpredictive there is an inherent tension between reflexive and volitional orienting, which de lays or abol ishes volitional orienting to the predicted location. A second possibil ity is that at an intermediate S O A both forms of orienting might be operating independently, and when performance at the two locations is contrasted there is no significant difference because performance is facilitated by attention at both locations. Experiment 1 tested these two ideas. In order to explore the first possibility, we extended the range of cue-target intervals to include long S O A s of 1200 and 1800 ms. W e reasoned that this would provide ample opportunity for attention to be oriented volitionally to the predicted (but not gazed-at) location, and that at these long S O A s any reflexive tendency to orient towards the gazed-at location should no longer be present. The second possibil ity was tested by increasing the number of target locations from two to four 1. In this way, on any given trial there would a lways be potential target locations 1 The addition of up and down target locations raises the possibil ity of introducing differential cuing effects, depending on whether orienting occurs on the horizontal axis or the vertical axis. In their study with real faces, Langton and Bruce (1999) found that although gaze cuing effects occurred only on the horizontal axis with inverted faces (Experiment 4), the effects were equivalent on both axes with upright faces (Exper iments 1 - 3). To rule out the possibil ity that there were axis effects in the present study, we conducted an A N O V A for each of our two experiments with the axis on which the target appeared (horizontal, vertical), S O A (105, 600, 1200, 1800 ms), and trial type (predicted, cued, N P - N C ) as within-subject factors. There was no interaction between axis and trial type for either gaze cues (F < 1, p >.95) or arrow cues (F < 1.7, p > .20). Ax is will therefore not be considered a factor in the present study. 78 that were neither gazed-at nor predicted. These locations would therefore provide a true basel ine against which to a s s e s s the al location of reflexive attention to the gazed-at location and volitional attention to the predicted location. The gaze direction of a centrally presented schemat ic face served as the cue, and target onset could occur to the left, to the right, above, or below the face. In order to a s s e s s any transitions from reflexive orienting to volitional orienting, we measured performance at a short S O A (105 ms) when reflexive orienting is typically observed, at an intermediate S O A (600 ms) when reflexive and volitional effects might both occur, and at two long S O A s (1200 and 1800 ms) when volitional attentional effects should predominate. Method Part icipants Twenty-four introductory psychology students (17 females and 7 males) reporting normal or corrected-to-normal vision participated in the experiment for course credit. All participants were unaware of the purposes of the experiment. Test ing was divided over two sess ions of less than one hour each , conducted on separate days . Apparatus and Stimuli The experiment was controlled by V S c o p e software (Rensink, 1995) on a 6100 Power Macintosh computer. Stimuli were presented on a 14-inch App le color monitor set to black and white and operating at a refresh rate of 66.7 Hz . R e s p o n s e time (RT) measures were based on keyboard responses. The face display, illustrated in Figure 5.1, consisted of a black line drawing of a face presented on a white background. The round face outline subtended 6.8°, and contained two circles representing the eyes , a smaller circle in the center of the monitor representing the nose and serving as the fixation point, and a straight line representing 79 the mouth. The eyes subtended 0.9°, and the center of each eye was located 1.0° to the left or right of the central vertical axis and 0.8° above the central horizontal axis. The nose subtended 0.2°. The mouth was 2.2° in length and was centered 1.3° below the nose. Black filled-in circles appeared within the eyes and represented the pupils. The pupils subtended 0.5°. For the left and right gaze direction cues , the pupils were centered vertically in the eyes , and were just touching either the left or right of the eyes ; for the up and down gaze directions, the pupils were centered horizontally in the eyes , and were just touching either the top or bottom of the eyes ; and for straight-ahead gaze , the pupils were centered both horizontally and vertically in the eyes . Note that when the pupils were presented, they appeared in a left, right, up, down, or straight gaze posit ion, and thus there was no motion artifact. The target st imulus demanding a detection response was a black capital letter F or T that measured 0.75° wide and 1.35° high, and was presented to the left, to the right, above or below the face. Target letters were centered on either the horizontal or vertical meridian, and the distance between central fixation (the nose) and the center of the target letter was 6.25°. Des ign Cue-target S O A (105, 600, 1200, or 1800 ms), gaze direction (left, right, up, down, or straight), and target identity (F or T) were selected randomly and with equal probability. W h e n gaze direction was left, right, up, or down, the target letter appeared at the location opposite to where the eyes were looking 7 5 % of the time (e.g., if the eyes looked up, the target was most likely to appear below the face). If a target did not appear at the predicted location, target location was selected randomly and with equal probability from among the three remaining alternative posit ions. W h e n gaze direction 80 was straight, a target was presented at one of the four locations (left, right, above, or below) randomly and with equal probability. There were four trial types: predicted trials, in which the target appeared at the predicted location (i.e., at the position opposite to the gazed-at location); cued trials, in which the target appeared at the gazed-at location; not predicted - not cued ( N P - N C ) trials, in which the target appeared at one of the two locations that were neither predicted nor gazed-at ; and nondirectional straight-gaze trials, in which the eyes looked straight ahead and the target could appear at any of the four target locations. Figure 5.2 illustrates the probabilit ies of the possib le target posit ions for predicted, cued , N P -N C trial types. E a c h of the two experimental sess ions was composed of 20 practice trials fol lowed by 12 blocks of 60 trials, for a total of 1440 test trials per participant. Approximately 8% of the test trials were catch trials randomly selected from the five gaze direction cues . Procedure The sequence of events on a target trial is illustrated in Figure 5. 1. Al l trials began with the presentation of a face with blank eyes . After 675 ms, pupils appeared within the eyes , looking left, right, up, down, or straight ahead . Then , after 105, 600, 1200 or 1800 ms, a target letter appeared to the left, to the right, above, or below the face. Both the gazing face and the target letter remained on the screen until a response was made or until 1500 ms had e lapsed, whichever came first. R T was measured from the time of target onset. The intertrial interval was 675 ms. Part icipants were seated approximately 57 cm from the monitor, and the experimenter ensured that they were centered with respect to the monitor and keyboard. They were told that each trial would begin with a line drawing of a face with 81 blank eyes , that pupils would appear in the eyes to create a face that was looking left, right, up, down, or straight ahead , and that after the appearance of the pupils, a capital letter (either F or T) would typically appear to the left, right, above, or below the face. Part icipants were instructed to press the spacebar with the index finger of their preferred hand as quickly as possible when a letter appeared on the sc reen . A l so , they were told that occasional ly there would be trials in which no target appeared , and that on these trials they should not respond, and just wait for the next trial to begin. The experimenter s t ressed that it was important to maintain fixation on the nose in the center of the face at all t imes. It is important to note that the identity of the target was irrelevant to the response task — subjects were merely required to detect target onset. Pas t research has demonstrated that when subjects are required to detect target onset they will normally not move their eyes to the target before making a response (Posner, 1980). Indeed it is difficult to get subjects to move their eyes before making a manual detection response because it s lows their R T performance significantly (Kingstone & Pratt, 1999). Thus we did not expect eye movements to be a confounding factor in our study. Moreover, we have confirmed e lsewhere that the standard attention effects produced by nonpredict ive gaze direction do not depend on eye movements of the participants (Fr iesen & Kingstone, 2003a). Never theless, we monitored the eye position of seven participants to be certain that this was the case . A s our results will show, our expectat ions were conf irmed: subjects who were eye-monitored rarely executed eye movements before responding, and their performance w a s indistinguishable from the performance of those who were not eye-monitored. Thus the ev idence indicates that eye movements were rarely occurring in our study and are thus not an explanatory concern for our data. 82 Before beginning each sess ion , participants were told that 7 5 % of the time the eyes looked left, right, up, or down, the target letter would appear at the location opposite to where the eyes were looking, and that when the eyes looked straight ahead , the target was equal ly likely to appear at any of the four possible target locations. Resul ts M e a n R T s , standard deviat ions, and error rates for Exper iment 1 are presented in Table 5.1. Anticipat ions (RTs < 100 ms), timed-out trials (no response) , R T s longer than 1000 ms, and incorrect responses (accidentally pressing a key other than the spacebar) , were classi f ied as errors and were exc luded from analys is. E a c h type of error accounted for less than 0 .5% of the target trials. The false alarm rate on catch trials was 1.02%. B e c a u s e these rates were so low, the error and false alarm data were not ana lyzed further. A n A N O V A was conducted with S O A (105, 600, 1200, and 1800 ms) and trial 2 type (predicted, cued , and basel ine) as within-subject factors . M e a n R T s for predicted, cued , and N P - N C trials at each S O A are illustrated in Figure 5.3. The A N O V A revealed that there was a significant main effect for S O A [F(3,23) = 66.66, p <0.0001], with R T becoming shorter as S O A lengthened (a standard foreperiod effect, Fr iesen & Kingstone, 1998; Mowrer, 1940). There was also a significant main effect for trial type [F(2,23) = 12.47, p < 0.0001], with R T being shortest on predicted trials, intermediate on cued trials, and longest on N P - N C trials. Finally, the S O A x trial type interaction was 2 Note that this A N O V A excludes straight-gaze trials and thus is not subject to any confounds inherent in compar ing performance across different cue condit ions (cf. Jon ides & Mack, 1984). In other words, for predicted, cued, and N P - N C trials, the cue condit ions prior to target onset are identical, i.e., averted gaze . In this way similarities and dif ferences between predicted, cued , and N P - N C can be attributed to attentional al location without being compromised by factors such as changes in attentional set or response strategies between averted- and straight-gaze condit ions. Per formance on straight-gaze trials ac ross both exper iments is considered in detail in the Genera l D iscuss ion . 83 significant [F(6,23) = 2.66, p < 0.02]. A n inspection of Figure 5.3 suggests that this interaction reflects changes in the cued condition and predicted condit ion relative to the N P - N C condit ion across S O A s . Specif ical ly, it appears that for the cued condit ion there was early facilitation that persisted steadily ac ross the two early S O A s (105 and 600 ms), and then d isappeared at the 1200 ms S O A . Converse ly for the predicted condit ion facilitation emerged first at the 600 ms S O A and persisted thereafter. P lanned t-tests (Bonferroni corrected, two-tailed) confirmed these observat ions. For the cued condit ion, performance was significantly faster than the N P - N C condit ion only at the 105 and 600 ms S O A s . Converse ly , for the predicted condit ion, performance facilitation relative to N P - N C was absent at the 105 ms S O A and present at the 600, 1200, and 1800 ms S O A s . Finally, the data for the seven eye-monitored participants revealed that the eye movement rate was low (2.7%). R T performance for these seven eye-monitored participants was compared with that of the first seven nonmonitored participants we tested, using a two-within ( S O A , trial type) one-between (monitored/nonmonitored) A N O V A . The interactions involving monitoring condition and trial type fell far short of signif icance (all F s < 1, all ps > 0.50). Thus , we are certain that eye movements were not involved in producing our cuing effects, and that the orienting we observed was covert in nature. D iscuss ion In the present counterpredict ive gaze experiment, the use of four possib le target locations made it possible to compare performance for targets appear ing at a location that was gazed-at but was unlikely to contain a target (cued trials) with R T s for targets appear ing at a location that was not gazed-at and yet equally unlikely to contain a target ( N P - N C trials). Similarly, our des ign al lowed us to compare R T s for targets appear ing 84 at a location that was not gazed-at but was likely to contain a target (predicted trials) with R T s for targets appear ing at a location that was not gazed-at and also not likely to contain a target ( N P - N C trials). At the 105 ms S O A , responses to targets occurring at a gazed-at and unlikely location were significantly faster than responses to targets occurring at a nongazed-at and unlikely location. This advantage for cued locations relative to the basel ine locations that were neither predicted nor cued persisted at the 600 ms S O A , and then d isappeared by 1200 ms S O A . These findings are consistent with the reflexive effects observed with nonpredictive gaze cues (e.g., Driver et a l . , 1999; Fr iesen & Kingstone, 1998; Langton & Bruce, 1999), and they are a lso consistent with the Driver et al . (1999, Experiment 3) finding that at a short S O A , gaze direction can produce a covert attention shift even when subjects have incentive based on cue predict iveness to shift attention to some other location. W e also observed clear ev idence that subjects can shift attention volitionally to a predicted location when it is not the gazed-at location: responses were reliably faster for predicted locations relative to locations that were neither predicted nor gazed-at at 600, 1200, and 1800 ms S O A . Our observat ion of this effect at the 600 ms S O A indicates that counterpredictive gaze does not delay volitional orienting, a possibil ity that was suggested by the results of Driver et al. 's counterpredictive gaze experiment (1999, Experiment 3). Rather, it suggests that in Driver et al. 's experiment, significant effects were not observed at 700 ms S O A because both reflexive and volitional orienting were occurring at that cue-target interval, with reflexive attention being directed to the gazed -at location and volitional attention being directed to the predicted location. A s a result, when these two locations were directly compared, there was no significant difference between them. Consistent with this notion, a direct compar ison between predicted and 85 cued trials in the present experiment revealed that the two were not statistically different at the 600 ms S O A [t (23) = 0.38 p > 0.70]. In sum, our data indicate that participants will orient attention to a gazed-at location even though a target is unlikely to appear there, and that they can a lso orient attention volitionally to a predicted, nongazed-at location. Thus , Experiment 1 replicates the Driver et al . (1999) finding that orienting to a gazed-at location is reflexive in the strong sense that it can occur even when participants are trying to direct their attention to a different location; and it adds the new finding that participants are indeed able to al locate attention volitionally to a location that is counter to the gazed-at location. In addition, it indicates why Driver et al . failed to observe a significant effect of volitional orienting with their counterpredict ive gaze cues . That is, it appears that reflexive orienting to a gazed-at location and volitional orienting to a predicted location opposite to the gazed-at location overlap in time. B e c a u s e in the Driver et al . study there were only two locations, facilitation effects at the gazed-at location and volitional orienting to the predicted location created the false impression that orienting was not occurring reliably at either location. Our study, which included basel ine locations that were neither predicted nor cued , indicates that precisely the opposite is the case ~ volitional orienting to the predicted location and reflexive orienting to the gazed-at location can both occur at an S O A (600 ms) that c losely approximates the 700 ms S O A at which Driver et al. 's null f inding was observed . Importantly, our finding that there was an R T advantage for both predicted and cued trials compared to N P - N C trials suggests that both forms of orienting might be operating concurrently. The design of the present experiment does not allow for a conclus ive demonstrat ion of s imul taneous orienting to two different locations because , necessar i ly , on each trial only one location was probed. However, a c loser examinat ion 86 of our data rules out the most plausible alternative explanat ions and favors our interpretation that our counterpredict ive gaze cues produced concurrent reflexive and volitional orienting. First, t-tests revealed that the reflexive cuing effect ( N P - N C minus cued) at 600 ms S O A was not statistically different from the reflexive effect at 105 ms S O A [t (23) = 0.54, p > 0.58], and that the volitional cuing effect ( N P - N C minus predicted) at 600 ms S O A was not statistically different from the volitional effect at 1200 ms S O A [t(23) = 1.51, p > 0.13]. This indicates that at the 600 ms S O A , R T benefits for the predicted target location are not replacing, or occurring at the expense of, R T benefits for the gazed-at target location. S e c o n d , one could argue that our finding of facilitation on both predicted and cued trials at the 600 ms S O A might reflect volitional orienting by roughly half of the subjects at the 600 ms S O A , and reflexive orienting by the other half of the subjects at the 600 ms S O A . The data at the 600 ms S O A , however, do not support this proposal : R T s for 15 participants were shorter on both predicted and cued trials than on N P - N C trials, significantly more than would be expected by chance a lone (x2< 0.0005). A third possible explanat ion for the R T advantage for both predicted and cued trials relative to N P - N C trials at the 600 ms S O A is that individual participants were "switching" between the two types of orienting, i.e., that they were shifting attention volitionally to the predicted location on some trials and shifting attention reflexively to the gazed-at location on other trials. If this were the case , then one would expect that the R T var iance of the predicted and cued distributions would be greater than the R T var iance of the N P - N C distribution. This is because on predicted and cued trials somet imes the target would appear at an attended location and somet imes the target would appear at an unattended location. In contrast, on N P - N C trials the var iance 87 would be lower because on every trial the target would appear at an unattended location. In other words, switching should result in greater var iance because it entails the combined distribution of two component distributions with different means (attended and unattended target locations) compared to the basel ine N P - N C condit ion which has only one component distribution (unattended target locations). W e tested this switching explanation by conduct ing an analys is of the within-subject standard deviat ions at the 600 ms S O A , and the results were clear-cut: average standard deviat ions were not different between predicted trials (81 ms) and N P - N C trials (78ms) [t=0.74, p>0.45], and indeed standard deviat ions were smal ler on cued (67 ms) trials than on N P - N C trials [t=2.81, p<0.01]. In agreement with this analysis, a visual inspection of individual subjects' R T distributions on predicted and cued trials at the 600 ms S O A revealed no evidence of bimodality. In sum, our ability to rule out these alternative explanat ions favors the conclus ion that reflexive orienting to the gazed-at location and volitional orienting to the predicted location can occur concurrently. Experiment 2 The results of Experiment 1 indicate that orienting to gaze direction is reflexive in a strong sense . That is, target detection is facilitated at a gazed-at location despite the fact that gaze direction predicts that a target is likely to appear at a different, nongazed-at location. A n d , most impressively, this facilitation occurs even when attention is being oriented volitionally to the predicted nongazed-at location. Al though it s e e m s reasonable to speculate that the results of Experiment 1 are unique to gaze direction, this position remains untested. Indeed, recent ev idence suggests that nonpredictive arrow cues can produce behavioral effects that look very similar to those produced by nonpredictive gaze cues (Ristic, Fr iesen & Kingstone, 2002; T ipples, 2002; for a d iscuss ion, see Kingstone, Smi lek, Rist ic, Fr iesen & Eas twood, 2003). The purpose of 88 Experiment 2 was to examine whether similar effects to those observed in Exper iment 1 would be observed with a counterpredict ive central arrow cue. Experiment 2 was identical in every way to Exper iment 1, except that an arrow served as the counterpredict ive directional cue. Two different arrows were used . One half of our participants v iewed a symmetr ical arrow cue, with an arrowhead at one end and a tail at the opposite end (e.g., <—<). The other half of our participants v iewed an asymmetr ical arrow cue, with an arrowhead at the leading end but with no tail (e.g., <--). This latter modification was made to examine whether gaze cues were strongly reflexive because they were perceptually weighted in the direction that they looked toward. For example, when the eyes are looking to the left, the pair of black pupils is not centered on the midline of the display, but is instead centered on some point slightly to the left of the midline. If this is an important factor, then the symmetr ic and asymmetr ic arrows should produce different effects on reflexive attention, i.e., the asymmetr ical arrow should produce stronger reflexive orienting. Method Part icipants Twenty-four introductory psychology students (20 females and 4 males) reporting normal or corrected-to-normal vision participated in the experiment for course credit. All participants were unaware of the purposes of the experiment, and none had participated in Experiment 1. Test ing was divided over two sess ions of less than one hour each , conducted on separate days . Eye-monitor ing was conducted as in Exper iment 1. Apparatus and Stimuli The apparatus used was identical to that used in Exper iment 1. Stimuli for Experiment 2 are illustrated in Figure 5. 4. The fixation display consisted of a black line drawing of a cross centered within a circle. The circle subtended 6.8° and was centered 89 in the middle of the monitor. The cross within the circle was composed of a horizontal line and a vertical line, each of which was 2.1° in length. The intersection of the two lines of the cross served as the fixation point. For half of the subjects, directional cues were provided by an arrow head and an arrow tail appear ing at either end of one of the two lines of the cross (i.e., 1.2° from central fixation, as measured from the intersection of the cross to the pointed end of the arrow head or tail); for the other half of the subjects, cues were provided by only an arrow head appear ing at one end of one of the two lines of the cross. The arrow heads (and tails) were composed of two lines 0.6° in length, and measured 0.8° high by 0.5° wide. A nondirectional cue (corresponding to the straight-gaze cue in Experiment 1) was provided by smal l l ines appear ing at the ends of both l ines of the cross, such that each arm of the cross ended in a perpendicular line measur ing 0.6°. The response stimuli and task were as in Experiment 1. Des ign and Procedure The experimental design and procedure were identical to those of Exper iment 1, with the except ion that the directional cues were arrows and the nondirectional cue was a cross with perpendicular l ines on the end of each arm. C u e to target S O A (105, 600, 1200, or 1800 ms), cue type (left, right, up, or down arrow, or nondirectional cross) , and target identity (F or T) were selected randomly and with equal probability. The probabilit ies of a target appear ing at any one of the four locations were the s a m e as in Experiment 1. W h e n the cue was an arrow pointing left, right, up, or down, the target letter appeared at the location opposite to where the arrow was pointing 7 5 % of the time and at one of the other three locations 2 5 % of the time; and when the cue w a s the nondirectional c ross , the target appeared with equal probability at any one of the four locations. Thus , there were four trial types with probabilit ies identical to those in 90 Experiment 1: predicted trials, in which the target appeared at the predicted location (i.e., at the position opposite to where the arrow was pointing); cued trials, in which the target appeared at the location toward which the arrow was pointing; not predicted - not cued ( N P - N C ) trials, in which the target appeared at one of the two locations that were neither predicted nor pointed at by the arrow; and nondirectional cross trials. Approximately 8% of the trials were catch trials randomly selected from the five cue types. A s was the case with Exper iment 1, each of the two experimental sess ions was composed of 20 practice trials fol lowed by 12 blocks of 60 trials, for a total of 1440 test trials per participant. Figure 5. 4 provides an illustration of the sequence of events on a test trial. Resul ts Mean R T s , standard deviat ions, and error rates for Experiment 2 are presented in Table 5.2. A s in Exper iment 1, anticipations, timed-out trials, R T s longer than 1000 ms, and incorrect responses were classi f ied as errors and were exc luded from analys is. E a c h type of error accounted for less than 0.4% of the target trials. The false alarm rate on catch trials was 1.38%. B e c a u s e these rates were so low, the error and false alarm data were not ana lyzed further. A n A N O V A was conducted with S O A (105, 600, 1200, and 1800 ms) and trial type (predicted, cued , and N P - N C ) as within-subject factors. Figure 5.5 illustrates R T s for predicted, cued, and N P - N C trials. A s in Experiment 1, there was a significant main effect for S O A [F(3,23) = 27.89, p <0.0001], reflecting a foreperiod effect, and there was a significant main effect for trial type [F(2,23) = 36.55, p < 0.0001], with R T s on predicted trials shorter overall than R T s on cued and N P - N C trials. The S O A x trial type interaction was a lso significant [F(6,23) = 5.23, p < 0.0001]. 91 P lanned t-tests (Bonferroni corrected, two-tailed) conducted as in Exper iment 1 revealed that the predicted condition was significantly faster than the N P - N C condition at all but the shortest S O A . This was precisely the s a m e result that was observed in Experiment 1. However, unlike Exper iment 1, the cued and N P - N C condit ions were statistically equivalent at all S O A s . Recal l that in Experiment 1 reflexive orienting was observed at the gazed-at location at both the 105 and 600 ms S O A s . A n A N O V A with S O A (105, 600, 1200, and 1800 ms) and trial type (predicted, cued, and N P - N C ) as within-subjects factors and with arrow type (symmetrical, asymmetr ical) as a between-subjects factor revealed that there were no significant effects involving arrow type [all Fs<2; all ps>0.16], confirming that our arrow effects were equivalent when the arrow cues may have had less directional sa l iency than our gaze cues (symmetrical arrows) and when the arrow cues were given greater directional weight (asymmetr ical arrows). A s in Experiment 1 the seven participants who were eye-monitored rarely produced eye movements (3.0%). And a compar ison of R T performance for these seven eye-monitored participants with that of the first seven nonmonitored participants produced no significant effects (all F s < 1.4, all ps > 0.20), once again confirming that eye movements were not involved in producing our cuing effects, and that the orienting we observed was covert in nature. D iscuss ion Exper iment 2 was identical to Experiment 1, with the except ion that arrows were used instead of gaze as the central ly-presented counterpredictive cue. The pattern of R T s for predicted but not cued target locations versus locations that were neither predicted nor cued was very similar to that obtained with counterpredict ive gaze in Exper iment 1; that is, a reliable advantage for targets occurr ing at the predicted location 92 was observed at 600, 1200, and 1800 ms S O A , indicating that participants were able to shift attention volitionally to the location where a target was likely to occur. However, the pattern of R T s for cued but not predicted target locations versus locat ions that were neither predicted nor cued was very different from that obtained in Exper iment 1: with arrows, there was never a significant advantage for targets occurr ing at the cued location. This difference between experiments was confirmed statistically, with an A N O V A compar ing gaze and arrows (cue type) between groups and reveal ing significantly different cuing effects (trial type) both as a function of S O A (cue type x trial type x S O A [F(9, 46) = 1.91, p < 0.05]), and when co l lapsed across S O A s (cue type x trial type [F(3,46) = 7.87, p < 0.0001]). Note that this difference between gaze and arrows cannot be attributed to a perceptual weighting toward the cued (gazed-at) location in Exper iment 1 because an asymmetr ical arrow in Exper iment 2 did not produce an advantage at the cued location, nor did it produce any difference from a symmetr ic arrow. G iven that nonpredictive arrows can produce relatively early facilitation (Ristic, Fr iesen & Kingstone, 2002), it may s e e m curious that in the present arrow cuing experiment the advantage for cued target locations did not reach signi f icance. This d iscrepancy may be due to dif ferences in design between the exper iments, such as the difference in the number of possible target locations (two in the Rist ic et al . study and four in the present study), or dif ferences in the distance between cued and uncued locations (180° in the Rist ic et al . study and 90° in the present study). However, in recent exper iments with nonpredictive arrows and four target locations (Ristic, Oik, Ho, & Kingstone, 2003), we observed early facilitation similar to that observed in the Rist ic et al . (2002) study with two target locations. Thus , we favor the more interesting and meaningful possibility that the d iscrepancy is due to differences in the predictive value 93 of the arrow cues across studies. In the Rist ic et al . (2002) study the arrow cues were spatially nonpredictive. In the present study they were spatially counterpredict ive. It appears then that the arrow cuing effect may be less strongly reflexive in nature than the gaze cuing effect, and may therefore be more vulnerable to observers ' top-down goals and expectat ions. If so , then it is reasonable that volitional orienting in a direction opposite to the arrow direction might undermine the reflexive orienting effect of the arrow stimuli in the present study. At any rate, our data indicate that while gaze and arrows are similar in their ability to produce a volitional shift in covert attention, arrow cues do not trigger a reflexive shift of attention to a location where a target is unlikely to appear. Genera l D iscuss ion Our counterpredictive gaze experiment (Experiment 1) replicated the finding of Driver et al . (1999) that subjects orient attention reflexively to a gazed-at location at a short S O A even though they expect the target not to appear there. This confirms that orienting to gaze direction is reflexive in a strong sense , i.e., that it can occur even against subjects' intentions. Exper iment 1 a lso demonstrated that subjects can direct attention volitionally to a nongazed-at location at longer S O A s . Moreover, the results of this experiment indicate that at an intermediate S O A , when both reflexive attention to the gazed-at location and voluntary attention to the likely location might be expected to occur, both may indeed have occurred. In other words, reflexive orienting to gazed-at locations and volitional orienting to likely locations exhibited different but over lapping t imecourses. This suggests that gaze-tr iggered orienting and volitional orienting might occur somewhat independently of one another, such that attention can be directed reflexively to one location and volitionally to another location at the s a m e time. Supplementary ana lyses supported this interpretation and failed to lend support to 94 alternative explanat ions, such as the possibil ity that our finding of reflexive and volitional co-occurrence was an artifact of averaging across different subjects or the possibil ity that it was due to subjects switching between one type of orienting on one trial and another type of orienting on another trial. The results of our second experiment with counterpredictive arrows suggested that the data pattern observed in Exper iment 1 may be unique to gaze direction cues . Exper iment 2 was identical in every way to Experiment 1 with the except ion that gaze direction cues were substituted with arrow cues . Yet the results were clearly very different. In both experiments, ev idence of covert voluntary orienting to the predicted target location was observed at 600, 1200, and 1800 ms S O A . However, in contrast to our f indings with counterpredict ive gaze cues, with counterpredictive arrow cues there was no ev idence of covert reflexive orienting to the cued location. The difference observed in the present study between gaze and arrow cues lends support to the notion that gaze direction may be a specia l attentional cue that can trigger reflexive shifts of attention that are in opposit ion to, and concurrent with, volitional shifts of attention. In the present study we chose to use a schemat ic face, rather than an image of a real face, to provide our gaze cue because such a s imple stimulus is more perceptually equivalent to other directional cues in the environment, such as the arrow cue we used . Nevertheless, one might wonder whether the gaze effects we observed with schemat ic faces can be general ized to more realistic looking faces and eyes . To our knowledge, only one study to date has directly compared the gaze cuing effects of schemat ic faces with those of real faces. Using schemat ic and real faces with various emotional express ions, Hietanen and Leppanen (in press) found that schemat ic faces produced similar, albeit somewhat larger, cuing effects to those produced by real faces at a 200 95 ms S O A . In numerous other studies, the reflexive gaze cuing effect has been observed with schemat ic faces (e.g., Fr iesen & Kingstone, 1998, 2003a, 2003b; Kingstone, Fr iesen, & G a z z a n i g a , 2000; Rist ic, Fr iesen, & Kingstone, 2002) and with real faces (e.g., Driver et al . , 1999; Hietanen, 1999; Langton & Bruce, 1999); and compar isons across studies suggest that in general the performance effects are equivalent. With regard more specif ical ly to the effects of counterpredictive gaze cues , Exper iment 1 of the present study with schemat ic faces replicates the findings produced by Driver et al . (1999, Exper iment 3) with real faces (i.e., there is reflexive orienting to the gazed-at but unlikely target location at a short S O A , and there is no difference between cued and predicted locations at an intermediate S O A ) . Our paradigm was des igned so that we could a s s e s s the effects of our directional cues by compar ing performance at cued and predicted locations with performance at locations that were neither predicted nor cued. The inclusion trials on which the target would appear at a location that was neither predicted nor cued ( N P - N C ) by the directional st imulus (i.e., gaze or arrow) provided the ideal basel ine for our purposes, because a target appear ing at one of those locations is preceded by exactly the same type of cue as a target appear ing at a cued location or a predicted location. A s was first noted by Jon ides and Mack (1984), failure to obtain such a basel ine measure leaves open the very real possibil ity that performance differences between cued and neutral trials (such as the straight-gaze and cross cues , in the case of the present study) may have nothing to do with attentional orienting and everything to do with one or more confounding factors such as arousal , effort, or strategy. By using basel ine trials that are directional cue trials we can make assessmen ts of reflexive and volitional orienting at different time windows with conf idence -- something that most studies have not been 96 able to do in the past (but see Kingstone & Kle in, 1991, and Danz iger & Kingstone, 1999, for two noteworthy except ions). In a previous study with nonpredictive gaze cues in which we treated straight-gaze trials as our neutral basel ine, we concluded that gaze direction cues produced benefits at gazed-at locations without any corresponding costs at nongazed-at locations (Fr iesen & Kingstone, 1998). The inclusion of similar "neutral" trials (straight-gaze trials in Experiment 1, and cross trials in Exper iment 2) in the design of the present study afforded us an opportunity to compare these neutral nondirectional cue trials (straight-gaze or cross) with our N P - N C directional cue trials. For each experiment, an A N O V A was conducted with trial type (directional N P - N C , nondirectional neutral) and S O A (105, 600, 1200, 1800 ms) as within-subject factors. For gaze cues (Experiment 1), there was a main effect for trial type, with R T 5 ms longer on N P - N C trials than on straight-gaze trials [F(1, 23)=11.11, p<0.005]; and the trial type x S O A interaction w a s not significant [F<1.0]. For arrow cues (Experiment 2), the main effect for trial type was not significant [F<1.0]; but the S O A x trial type interaction was significant [F (3, 23)=7.11, p<0.0005]. Inspection of the data suggested that this interaction was caused by shorter R T s on N P - N C trials than on central c ross trials at the 105 ms S O A . In agreement with this interpretation, when the 105 ms S O A trials were removed from the A N O V A , a completely different result was obtained. Now, there was a marginally significant main effect for cue type, with R T 3 ms longer on N P - N C trials than on cross trials [F(1, 23)=3.11, p<0.10], and the trial type x S O A interaction fell far short of s igni f icance [F<1.6]. The overall pattern of results with the nondirectional trials converges with the results we reported using directional trials as our basel ine, i.e., that responses are primarily facilitated at cued and/or predicted target locations. A s for the one anomalous 97 finding just d iscussed (shorter R T on N P - N C arrow trials than on central c ross trials at 105 ms S O A ) , it provides an illustration of the inherent danger of failing to include an appropriate basel ine measure at the time that attention is cued (Jonides and Mack, 1984). If we had not included directional basel ine N P - N C trials in our experiment, and if we had compared predicted and cued trials to the nondirectional c ross trials, we would have been misled into thinking that there was early facilitation on both predicted trials and cued trials. But our data reveal that directional basel ine trials are a lso "facilitated" relative to neutral. Clear ly, there really is no cuing effect occurr ing at the 105 ms S O A (predicted and cued trials are not significantly faster than N P - N C trials), and the "neutral" nondirectional st imulus (i.e., the cross) is being treated differently from the directional arrow cues at this early S O A . Thus, it is important to note that although nondirectional neutral cues (such as our straight gaze and cross cues) might general ly serve as a reasonable basel ine, they do not a lways do so. The different but over lapping t imecourses of reflexive orienting to a gazed-at location and volitional orienting to a likely target location observed in Exper iment 1 suggests that the two forms of orienting may be independent, and thus that they may be subserved by different attentional sys tems or subsys tems. There is considerable ev idence in the attentional literature indicating that reflexive orienting to a sudden onset at a peripheral location and volitional orienting to an expected target location occur by way of different brain pathways. Ref lexive orienting to a sudden onset in the periphery is thought to involve the superior col l iculus (SC) , working in concert with parietal cortex (Rafal , Henik, & Smith, 1991; Rafal , Posner , Fr iedman, Inhoff, & Bernstein, 1988), whereas volitional orienting to an expected target location is thought to involve frontal and parietal areas (Corbetta, Miez in , Shu lman, & Petersen, 1993; Posne r 1995; Posne r 98 and Raich le , 1994). It s e e m s likely, however, that reflexive orienting triggered by gaze direction does not occur by way of either of these pathways. Severa l l ines of ev idence suggest that gaze-tr iggered orienting does not occur by way of the subcort ical pathway. First, in their study with split-brain patients, Kingstone, Fr iesen, and G a z z a n i g a (2000) demonstrated that reflexive orienting to gaze direction is lateralized to one cortical hemisphere. S e c o n d , in a recent eye movement study, Fr iesen and Kingstone (2003a) found that gaze direction cues did not activate or predisengage the oculomotor sys tem, suggest ing that orienting to gaze direction does not engage the S C . And third, Fr iesen and Kingstone (2003b) demonstrated that reflexive orienting to gaze direction can co-occur with IOR (which is subserved by the S C ) . Similarly, the finding of the present study that reflexive orienting to a gazed-at location and volitional orienting to a different location might co-occur suggests that attention to gaze does not occur by way of the frontal-parietal pathway that underl ies volitional orienting. This conclus ion is consistent with three other results suggest ing that gaze-tr iggered orienting is not simply a wel l- learned form of volitional orienting. First, Rist ic, Fr iesen, and Kingstone (2002) found that preschool chi ldren showed greater orienting effects than adults in response to nonpredictive gaze direction cues , despite the fact that young children are thought to be poor at volitional orienting (Brodeur, Trick, and Enns , 1997). S e c o n d , Hood, Wi l len, and Driver (1998) found that infants were faster to make s a c c a d e s to peripheral targets that were cued nonpredictively by the gaze direction of a central face, and conc luded that gaze-tr iggered orienting is in place very early in development (but see Farroni, Johnson , Brockbank, & Simion (2000) for an alternative explanation). And third, in their split-brain patient study, Kingstone, Fr iesen, and G a z z a n i g a (2000) found that although only the cortical hemisphere specia l ized for 99 face and gaze processing oriented reflexively in response to nonpredict ive gaze cues , both hemispheres oriented volitionally in response to predictive gaze cues . S o what might the gaze-tr iggered reflexive attention pathway be? Kingstone, Fr iesen, and G a z z a n i g a (2000) proposed that orienting to gaze direction might be subserved by a temporal-parietal pathway, with cel ls in inferotemporal cortex (IT) processing face and eye information, cells in the superior temporal su lcus (STS) processing the direction of gaze , and cells in parietal cortex shifting attention to the gazed-at location. E a c h of these brain regions has s ince been implicated in gaze direction processing in a number of human neuroimaging studies (e.g., Hoffman & Haxby, 2000; Kato et al . , 2001; P u c e , Al l ison, Bent in, Gore , & McCar thy , 1998; Wicker , Michel , Henaff, & Decety, 1998). Note that all three of the attentional pathways d iscussed here — the subcort ical reflexive pathway, the cortical volitional pathway, and the proposed cortical gaze direction pathway — involve parietal cortex. How, then, could attention be shifted reflexively to a gazed-at location and volitionally to a different location at the same t ime? One possibil ity is that volitional inputs from frontal cortex and gaze inputs from temporal cortex activate different parietal neurons. In a recent fMRI study that compared peripheral target detection versus volitional orienting, Corbetta, K incade, Oll inger, McAvoy , and Schu lman (2000) found ev idence for this type of dissociat ion, with temperoparietal cortex activated during target detection, and intraparietal cortex activated during volitional orienting. To our knowledge, the present study is the first to demonstrate that gaze cues and arrow cues can produce qualitatively different behavioral results in intact observers. In their recent study with nonpredict ive gaze and arrow cues, Rist ic, Fr iesen, and Kingstone (2002) found that nonpredictive gaze cues and nonpredict ive arrow cues produced similar R T patterns in normal participants (both adults and children). 100 Differences in the effects of the two types of directional cue were revealed only when the performance of a split-brain patient was examined: nonpredictive arrow cues triggered orienting in both hemispheres, whereas in a previous study of the s a m e patient (Kingstone, Fr iesen, and G a z z a n i g a , 2000) nonpredict ive gaze cues triggered orienting only in the hemisphere specia l ized for face process ing. Based on this difference in lateralization for the two cue types, Rist ic, Fr iesen, and Kingstone concluded that gaze is indeed spec ia l . The present study, however, demonstrates that apart from the issue of lateralization of face processing, gaze and arrow cues can trigger qualitatively different behavioral effects. W h e n each of these directional cues is put into competit ion with volitional orienting, orienting to gaze direction persists, whereas orienting to arrows is abol ished. In sum, the results of the present study confirm that attentional orienting toward a gazed-at location is reflexive, not only in the sense that it occurs when participants do not have any incentive to attend to the gazed-at location (as is the case in nonpredict ive gaze experiments), but a lso in the stronger sense that it can occur even when participants are attending volitionally to an opposite location. Our finding that reflexive and volitional orienting in response to gaze direction appear to co-occur suggests that the two may be subserved by distinct and separable mechan isms. Arrow cues can also produce reflexive shifts of attention (Ristic, Fr iesen & Kingstone, 2002; T ipp les, 2002), but unlike eyes they do not do so when they are counterpredict ive. The implication is that while many directional cues might trigger reflexive shifts of attention when they are spatially nonpredictive, they are not all equal . In particular, gaze cues appear to be more strongly reflexive than arrow cues , very possibly because they a c c e s s a neural architecture that is specia l ized for processing eye direction. Table 5.1 101 Condit ion M S D % E 105 ms S O A 600 ms S O A 1200 ms S O A 1800 ms S O A Predicted 390 55 0.94 Cued 385 51 0.93 N P - N C 397 58 0.85 Straight-gaze 396 56 0.99 Predicted 355 53 0.92 C u e d 356 46 0.58 N P - N C 367 54 1.69 Straight-gaze 360 51 0.68 Predicted 338 50 0.48 C u e d 354 55 0.56 N P - N C 356 49 0.93 Straight-gaze 347 46 0.69 Predicted 346 46 0.55 C u e d 349 50 0.76 N P - N C 357 47 0.38 Straight-gaze 353 46 0.25 Table 5.1. Mean R T s (in ms), Standard Deviat ions, and Errors Rates (%) for Experiment 1. Note. N = 24. Error rates represent the percentage of test trials from each cell exc luded as anticipations, key press select ion errors, t imed-out trials, or trials with R T > 1000 ms. S O A = stimulus onset asynchrony. 102 Table 5.2 Condit ion M S D % E 105 ms S O A 600 ms S O A 1200 ms S O A 1800 ms S O A Predicted 379 53 0.69 C u e d 377 56 0.97 N P - N C 385 52 0.48 C r o s s 398 54 0.83 Predicted 348 52 0.78 C u e d 374 61 0.19 N P - N C 372 50 0.76 C r o s s 366 46 0.44 Predicted 333 49 0.66 C u e d 355 50 0.74 N P - N C 353 45 0.56 C r o s s 349 46 0.56 Predicted 337 47 0.81 C u e d 358 50 0.19 N P - N C 351 47 0.65 C r o s s 352 48 0.75 Table 5.2. Mean R T s (in ms), Standard Deviat ions, and Errors Rates (%) for Exper iment 2. Note. N = 24. Error rates represent the percentage of test trials from each cell exc luded as anticipations, key press select ion errors, t imed-out trials, or trials with R T > 1000 ms. S O A = stimulus onset asynchrony. Figure 5.1. 103 time Fixation Display Gaze Cue Display Target Display 675 ms 1 0 5 , 6 0 0 , 1 2 0 0 , un t i l response, or 1800 ms or 1500 ms Figure 5.1. Illustration of the trial sequence in Experiment 1. Each trial began with the presentation of a face with blank eyes. After 675 ms, pupils appeared in the eyes, looking left, right, up, down, or straight ahead (the gaze cue). Then, after 105, 600, 1200, or 1800 milliseconds (ms), the letter F or T (the target) appeared to the left or to the right, above, or below the face. The target was likely to appear at the location opposite to the gazed-at location 75% of the time the eyes looked left, right, up, or down. 104 Figure 5.2. N P - C 8% P - N C 7 5 % Figure 5.2. Illustration of the three trial types that were possible when gaze was directed at one of the four target locations in Experiment 1. Predicted = target occurs at the predicted (not cued) location. Cued = target occurs at the cued (not predicted) location. NP-NC = target occurs at a location that is neither predicted nor cued. Numbers represent the percent probability (rounded to the nearest percentage point) of the target's appearance at each location. 105 Figure 5.3. G a z e C u e 3 3 ( H 105 600 1200 1800 S O A Figure 5.3. Experiment 1 mean RTs for counterpredictive gaze cues as a function of cue-target stimulus onset asynchrony (SOA) and trial type. Predicted = target occurs at the predicted (not cued) location. Cued = target occurs at the cued (not predicted) location. N P - N C = target occurs at a location that is neither predicted nor cued. 106 Figure 5.4. A. Symmetrical Arrow Cue time Fixation Display 675 ms Arrow Cue Display 105, 600, 1200, or 1800 ms Target Display until response, or 1500 ms B. Asymmetrical Arrow Cue Figure 5.4. Illustration of the trial sequence in Experiment 2. Each trial began with a cross at central fixation. After 675 ms, an arrow with a head and a tail (A) or an arrow with only a head (B) appeared on one of the two lines of the cross, creating an arrow pointing left, right, up, or down. On nondirectional cross trials, small perpendicular lines appeared at the ends of the lines of the cross. Then, after 105, 600, 1,200, or 1,800 ms, a target letter (F or T) appeared to the left of, to the right of, above, or below the cross. Trial types and probabilities were the same as those for counterpredictive gaze direction cues (see Figure 6.2). 107 Figure 5.5. A r r o w C u e 400-1 390-330H 105 600 1200 1800 S O A Figure 5.5. Experiment 2 mean RTs for counterpredictive arrow cues as a function of cue-target stimulus onset asynchrony (SOA) and trial type. Predicted = target occurs at the predicted (not cued) location. Cued= target occurs at the cued (not predicted) location. NP-NC = target occurs at a location that is neither predicted nor cued. 108 References Brodeur, A . , D., Trick, M. , I, and Enns , J . T. (1997). Select ive attention over the l i fespan. In Burack A . J . & Enns J . T. (Eds.). Attention, Development, and Psychopatho logy (pp. 74-97). New York: Guilford P ress . C h e a l , M . L , & Lyon, D. R. (1991). Central and peripheral precuing of forced-choice discrimination. Quarterly Journal of Experimental Psycho logy , 43A, 859-880. Corbetta, M. , K incade, J . M. , Oll inger, J . M. , McAvoy , M. P., & Schu lman , G . L. (2000). Voluntary orienting is d issociated from target detection in human posterior parietal cortex. Nature Neurosc ience, 3 A 292-297. Corbetta, M. , Miez in , F. M. , Shu lman, G . L., & Petersen, S . E. (1993). A P E T study of v isuospat ial attention. Journal of Neurosc ience, 13, 1202-1226. Danziger, S . , & Kingstone, A . (1999). Unmask ing the inhibition of return phenomenon. Percept ion and Psychophys ics , 61 , 1024-1037. Driver, J . , Davis, G . , Ricciardel l i , P. , Kidd, P., Maxwel l , E. , & Baron -Cohen , S . (1999). G a z e perception triggers visuospat ial orienting by adults in a reflexive manner. V isua l Cogni t ion, 6, 509-540. Farroni, T., Johnson , M. H., Brockbank, M., & S imion, F. (2000). Infants use of gaze direction to cue attention: The importance of perceived motion. V isua l Cogni t ion, 7 ,705 -718 . Fr iesen, C . K., & Kingstone, A . (1998). The eyes have it!: Ref lexive orienting is triggered by nonpredict ive gaze . Psychonomic Bulletin and Review, 5, 490-495. Fr iesen, C . K., & Kingstone, A . (2003a). Covert and overt orienting to gaze direction cues and the effects of fixation offset. NeuroReport , 14 A 489-493. Fr iesen, C . K., & Kingstone, A . (2003b). Abrupt onsets and gaze direction cues trigger independent reflexive attentional effects. Cognit ion, 87x B1 -B10 . 109 Hietanen, J . K. (1999). Does your gaze direction and head orientation shift my visual attention? NeuroReport , 10 ,3443-3447 . Hietanen, J . K. & Leppanen , J . M. (2003). Does facial express ion affect attention orienting by gaze direction c u e s ? Journal of Exper imental Psycho logy: Human Percept ion and Per formance, 29. 1228-1243. Hoffman, E. A . & Haxby, J . V . (2000). Distinct representat ions of eye gaze and identity in the distributed human neural sys tem for face perception. Nature Neurosc ience, 3, 80 - 84. Hood, B. M., Wi l len, J . D., & Driver, J . (1998). Adults ' eyes trigger shifts of visual attention in human infants. Psycholog ica l Sc ience , 9, 131-134. Jon ides , J . (1981). Voluntary versus automatic control over the mind's eye 's movement. In J . B. Long and A . D. Badde ley (Eds.) , Attention and Per formance IX (pp. 187-203). Hi l lsdale, N J : Er lbaum. Jon ides , J . & Mack, R. (1984). On the cost and benefit of cost and benefit. Psycho log ica l Bulletin. 96 A 29-44. Kato, C , Matsuo, K., Ma tsuzawa, M., Mor iya, T., Glover , G . H., & Nakai , T. (2001). Activation during endogenous orienting of visual attention using symbol ic pointers in the human parietal and frontal cort ices: a functional magnet ic resonance imaging study. Neurosc ience Letters. 314 A 5-8. Kingstone, A . , Fr iesen, C . K., & G a z z a n i g a , M. S . (2000). Ref lexive joint attention depends on lateral ized cortical connect ions. Psycholog ica l Sc ience , 11, 159-165. Kingstone, A . , & Klein, R. (1991). Combin ing shape and position expectanc ies: Hierarchical process ing and select ive inhibition. Journal of Exper imental Psycho logy: Human Percept ion & Per formance, 17, 512-519. Kingstone, A . & Pratt, J . (1999). Inhibition of return is composed of attentional and oculomotor p rocesses . Percept ion & Psvchophvs ics , 61 . 1046-1054. Kingstone, A . , Smi lek, D., Rist ic, J . , Fr iesen, C . K. & Eas twood, J . D. (2003). Attention, researchers! It is time to take a look at the real world. Current Directions in Psycholog ica l Sc ience , 12, 176-180. Klein, R. M. (2000). Inhibition of return. Trends in Cognit ive Sc iences , 4, 138-147. Klein, R. M., Kingstone, A . & Pontefract, A . (1992). Orienting of visual attention. In K. Rayner (Ed.), Eve Movements and V isua l Cogni t ion: S c e n e Percept ion and Read ing , (pp. 46-63). North-Hol land: Elsevier Sc ience Publ ishers B. V . Langton, S . R. H., & Bruce, V . (1999). Ref lexive socia l orienting. V isua l Cogni t ion, 6, 541-567. Langton, S . R. H., Watt, R. J . , & Bruce, V . (2000). Do the eyes have it? C u e s to the direction of socia l attention. Trends in Cognit ive Sc iences , 4, 50-59. Mowrer, O . H. (1940). Preparatory set (Expectancy) - S o m e methods of measurements . Psycholog ica l Rev iew Monograph, 52 (Whole No. 233). Muller, H. J . , & Rabbitt, P. M. A . (1989). Ref lexive and voluntary orienting of visual attention: time course of activation and resistance to interruption. Journal of Experimental Psycho logy: Human Percept ion and Per formance, 15, 315-330. Posner , M. I. (1980). Orienting of attention. Quarterly Journal of Exper imental Psycho logy. 32, 3-25. Posner , M. I. (1995). Attention in cognitive neurosc ience: A n overview. In M. S . G a z z a n i g a (Ed.), The Cognit ive Neurosc iences (pp. 615-624). Cambr idge, M A : MIT P ress . Posner , M. I., & C o h e n , Y . (1984). Components of V isua l Orienting. In H. Bouma & D. G . Bouwhuis (Eds.) , Attention and Per formance X (pp. 531-556). Hi l lsdale, N J : Er lbaum. Posner , M. I., C o h e n , Y . , & Rafa l , R. D. (1982). Neural sys tems control of spatial orienting. Proceed ings of the Roya l Society of London, 298, 187-198. Posner , M. I., Rafa l , R. D., Choate , L. S . , & Vaughan , J . (1985). Inhibition of return: Neural basis and function. Cognit ive Neuropsychology, 2, 211-228. Posner , M. I., & Raich le , M. E. (1994). Images of Mind. New York: Scientif ic Amer ican Library. Puce , A . , Al l ison, T., Bent in, S . , Gore , J . C , & McCar thy , G . (1998). Tempora l cortex activation in humans viewing eye and mouth movements. The Journal of Neurosc ience, 18, 2188-2199. Rafa l , R., Henik, A . , & Smith, J . (1991). Extrageniculate contributions to reflex visual orienting in normal humans: A temporal hemifield advantage. Journal of Cognit ive Neurosc ience, 3, 322-328. Rafa l , R. D., Posner , M. I., Fr iedman, J . F., Inhoff, A . W. , & Bernstein, E. (1988). Orienting of v isual attention in progressive supranuclear palsy. Brain, 111, 267-280. Rensink, R. A . , V S c o p e 1.2.7, M ic roPsych Software, Vancouver , British Co lumb ia , C a n a d a , copyright 1991 - 1995 Rist ic, J . , Fr iesen, C . K., & Kingstone, A . (2002). A re eyes spec ia l? It depends on how you look at it. Psvchonomic Bulletin and Rev iew. 9, 507-513. Rist ic, J . , Oik, B., Ho, S . , & Kingstone, A (2003). Endogenous orienting: What have we been measur ing? Poster presented at the 10 t h Annua l Meet ing of the Cognit ive Neurosc ience Society, New York, N Y , April 2003. 112 Tipples, J . (2002). Eye gaze is not unique: Automat ic orienting in response to uninformative arrows. Psychonomic Bulletin & Rev iew, 9, 314-318. Wicker , B. F., Michel , F., Henaff, M. , & Decety, J . (1998). Brain regions involved in the perception of gaze : A P E T study. Neuro image, 8, 221-227. 113 C H A P T E R 6 Attentional Control and Ref lexive Orienting to G a z e and Arrow cues A version of this chapter has been submitted for publication to Psychonomic Bulletin & Review. Rist ic, J . , Wright, A , & Kingstone, A . Attentional control and reflexive orienting to gaze and arrow cues . 114 In everyday life human eyes are a vital nonverbal socia l cue that enables fluent social communicat ion between individuals. Transmiss ion of gaze information is facilitated by a uniquely human eye morphology, with its dark iris surrounded by a high contrast white sc lera, that al lows people to accurately convey to others where they are looking (Kobayashi & Kosh ima , 1997). Percept ion of gaze information is a lso supported by a neural architecture, the superior temporal su lcus (STS) that is dedicated to the processing gaze information (Al l ison, P u c e & McCar thy , 2000). S igns of eye gaze communicat ion is evident shortly after birth, with infants as young as 2 or 3 days preferentially looking toward the eyes of another face (Batki et a l , 2000; Farroni et a l , 2002) and by the end of the first year most infants will look reliably toward where someone e lse is looking (Hood, Wil len & Driver, 1998). Researchers have deve loped a simple task to measure the tendency for children and adults to attend to where someone e lse is looking (Hood, Wil len & Driver, 1998; Fr iesen & Kingstone, 1998). Fr iesen and Kingstone (1998) asked adult participants to detect, local ize or identify visual targets that appeared to the left or right of a central face on a computer monitor. The key manipulation was that just before a target appeared , the eyes in the face on the computer sc reen would look to the left or right of center. Fr iesen and Kingstone found that the response time (RT) to a visual target was shortest when it appeared at the gazed-at location.. This R T advantage for a target at the cued (gazed-at) location occurred despite the fact that the participants knew in advance that the eyes in the face did not predict where a target would occur. Based on these data, and the fact that the R T advantage at the cued location emerged very rapidly, Fr iesen and Kingstone suggested that the laboratory paradigm engaged brain mechan isms such as the S T S that are dedicated to processing and orienting toward 115 where other people are looking (see a lso Driver et a l . 1999; Langton, Watt & Bruce, 2000). Subsequent to these original investigations, a body of ev idence emerged showing very clearly that other directional cues , such as f ingers, words, and arrows will produce an attention effect that is comparable to the attention effects produced by eyes (Gibson and Kingstone, in press; Hommel et al , 2001; Tipples, 2002; Watanabe , 2002). Col lect ively, these findings bring into question the original idea that the eye gaze paradigm descr ibed above is tapping into mechan isms that are specif ic to socia l orienting to eyes . Fr iesen, Rist ic and Kingstone (2004) addressed this issue directly by testing whether arrows trigger attention shifts that are as strongly reflexive as eyes . They compared performance elicited by eyes and arrows when each of these cues were counterpredict ive, e.g. , if a participant was shown eyes or arrows indicating a left location, the target was highly likely to occur at the opposite right location. Fr iesen et a l . , found that only eyes triggered an initial, rapid reflexive shift in attention to the cued (gazed-at) location. In other words, participants were unable to avoid attending reflexively to where the eyes were looking, but they were able to avoid attending reflexively to where arrows were pointing. This finding — that eyes are more strongly reflexive than arrows — lends support to the idea that the eye gaze paradigm is tapping into mechan isms that are speci f ic to social orienting to eyes. W h e n these data are considered as a whole, it appears that there is a wide range of stimuli that can produce reflexive shifts in spatial attention. S o m e central cues , like gaze direction, trigger rapid reflexive shifts of attention that are strongly reflexive and hence they are relatively insensit ive to top-down inhibitory control (Fr iesen, Rist ic & Kingstone, 2004; Rist ic & Kingstone, 2005). At the other end of the extreme are stimuli 116 like numbers, that can produce shifts of attention when they are spatially nonpredictive (Fischer et al . 2003) but this orienting effect is both slow to emerge and extremely sensit ive to changes in the top-down mental set adopted by a participant (Ristic, Wright & Kingstone, in press; Gal fano, Ruscon i & Umilta, in press). Arrow stimuli would appear to fall somewhere in the middle, producing rapid shifts of attention even when they are known to be spatially nonpredictive, but their effects are relatively amenab le to top-down control. A recent study by Pratt and Hommel (2003) provides an excel lent illustration of arrow cues ' sensitivity to top-down control. Pratt and Hommel found that when participants are set to respond to a target stimulus of a particular color, a task irrelevant arrow st imulus that shares the target color will trigger a shift in spatial attention. For example, if the set is for a target that is colored green, attention is shifted to the location cued by a green arrow; and if the set target is for a target that is colored blue, attention is shifted to the location cued by a blue arrow. In the present study we asked whether these arbitrary cue-target congruency effects found for arrows will a lso occur for eyes . B a s e d on our hypothesis that the attention effects for arrows are more amenable to arbitrary top-down assoc ia t ions than eyes , the strong prediction is that arrows will be significantly more sensit ive to cue-target color cont ingencies than eyes . Thus, we predict that arrows will produce cuing effects that are specif ic to congruent cue-target color cont ingencies (Pratt and Hommel , 2003). In contrast, because eyes are more strongly reflexive their attention effect may extend ac ross congruent and incongruent cue-target color cont ingencies. Finally, it is perhaps noteworthy that this is the first time that the effects of nonpredictive eyes and arorws are being tested directly against a benchmark effect establ ished by nonpredictive arrows. To date compar isons between the reflexive 117 orienting effects of eyes and arrows have consistently examined whether arrows can produce an outcome that is comparable to the effect of eyes (e.g., T ipples, 2002; Ristic, Fr iesen & Kingstone, 2002; Bay l iss , di Pel legr ino & Tipper, 2005). In the present study the fundamental nature of this compar ison is reversed, and hence the conceptual framework and quest ions that emerge are now being driven by the effects elicited by nonpredictive arrow stimuli. Experiment 1 In Pratt and Hommel 's (2003) original investigation participants were given a color precue that defined the response target, e.g., if the color precue was a blue patch then participants were to press a key if the target was colored blue and to refrain from responding if the target was not colored blue. Prior to a target appear ing, four arrows appeared at central fixation, each colored differently and each pointing in different directions (left, right, up, down). The key finding was that R T was fastest when the colored target appeared at the location cued by a matching colored arrow (e.g., a blue target appear ing at the location cued by a blue arrow). Importantly, this spatial color-target cont ingency effect occurred despite the fact that the location of the target was not predicted by the color or direction of the color-congruent arrow. The aim of the present experiment is to establ ish that this reflexive spatial cue-target color cont ingency effect can be replicated in a simpler task, e.g., in a task that has no color precue and only one arrow cue. In Experiment 1 participants were asked to press one key if the target was white and another key if the target was black. Prior to target onset, a spatially irrelevant arrow cue, colored white or black, appeared at central fixation and pointed left or right. The color and direction of the arrow cue did not predict the color or location of the target. The question was whether a reflexive spatial effect of 118 the arrow cue would be greater for a congruently colored target compared to an incongruently colored target, as suggested by Pratt and Hommel (2003). Method Part icipants Forty participants were ass igned randomly to two different groups: one group received a white arrow cue, and one group received a black arrow cue. Apparatus & Stimuli All stimuli were black and white line drawings presented on a gray background. The stimuli were shown on a P o w e r P C Macintosh computer connected to a 15-in color monitor set to black and white. The central arrow cue was a triangle measur ing 1° of visual angle at its base and 0.7° in height. The target (1° in s ize) was square- or d iamond-shaped. The arrow cue was presented at the center of the screen and targets appeared peripherally 3° away from center fixation. R e s p o n s e keys were " z " and "/" on a computer keyboard, with the left index finger pressing the " z " key and the right index finger pressing the "/" key. Color assignment to the response keys was counterbalanced between participants. Design The stimuli and sample sequence of events are illustrated in Figure 6.1. The black arrow and white arrow groups each performed a color discrimination task, pressing one key if the target was black and the other key if the target was white. E a c h group performed 480 color discrimination trials. For both black arrow and white arrow groups, the central arrow cue indicated one of the two possible directions (left or right), and the target could appear either on the left, right, up or down. Th is created two different cue validity condit ions: cued trials, in which the target occurred either on the left or right s ide as indicated by the arrow 119 (p=.25); and uncued trials in which the target occurred at the location that was not cued by the arrow (uncued locations are co l lapsed as an analys is of var iance revealed that this factor had no effect (F<1) on performance). C u e direction, target location, and target features (color and shape) were selected randomly and presented with equal probability. R T to press a key was measured in mi l l iseconds (ms) and t imed from target onset. Procedure Al l trials began with a 105 ms presentation of a central arrow cue pointing left or right. Fol lowing a 210 ms st imulus onset asynchrony ( S O A ) a target appeared left, right, above, or below the cue. The trial was terminated on response or after 1005 ms whichever occurred first. The intertrial interval was 2505 ms. Part icipants were seated centered in front of the computer sc reen at an approximate distance of 57 cm . It was emphas ized that the direction of the arrow and its color did not predict the location, color, or shape of the target. Al l participants were asked to respond as quickly and as accurately as possib le and to maintain central fixation throughout the experiment. Resul ts Anticipations (RT< 100 ms), timed-out responses (RT> 1000ms), and incorrect key presses were classif ied as errors and exc luded. These errors occurred on only 5.2% of the trials. Most importantly, as the subsequent ana lyses are conducted on correct RT , in no condition was the R T data contradicted by a speed-accuracy tradeoff, e.g. , faster R T at the cued location than the uncued location accompan ied by more errors at the cued location than the uncued location. The mean R T s for cued and uncued targets are presented in Figure 6.2A as a function of cue-target color congruency. Note that a spatial attention effect (cued R T < 120 uncued RT) is greater for color congruent cue-target stimuli than incongruent cue-target stimuli. That is, for the black arrow group R T is faster for the cued location versus the noncued location when the target is black; and for the white arrow group R T is faster for the cued location versus the noncued location when the target is white. Separate A N O V A s for each group confirmed this observat ion, with cue validity (cued/uncued) and target color (black/white), interacting [Black arrow group F(1,19)=4.7,p<.05; White arrow group F(1,19)=4.8,p<.05]. The main effects of target color were a lso reliable [Black arrow F (1, 19)= 24.6, p<0001 ; White arrow F(1,19)=27.0,p<0001], indicating that for each group R T is faster overall for the target color that is congruent with the color of the arrow cue. No other effects were significant [all Fs<2.4, all ps>.1]. The above ana lyses strongly suggest that the cue x validity interaction var ies with group, and this was confirmed by an A N O V A [F(1,38)=9.0,p<.01]. Group also interacted with target color, [F (1,38)=51.5,p<.0001], agreeing with the above observat ion that R T is faster overall when the color of the target is congruent with the color of the cue. No other effects were significant [all Fs<3.5,ps>.05]. D iscuss ion The data from this experiment show that the reflexive attention shift that is triggered by a nonpredict ive arrow cue produces a performance benefit that is specif ic to color congruent targets. This result provides an important conceptual replication and extension to the Pratt and Hommel (2003) study, as it demonstrates that an arbitrary cue-target color cont ingency effect extends to the simple situation of one arrow cue presented in a single color. The key quest ion now is whether the reflexive attention effect that is elicited by eye direction will a lso demonstrate a cue-target color cont ingency effect. A s outlined in the introduction, there are good reasons to think that eyes will be more strongly reflexive 121 and hence less sensit ive than arrows to arbitrary cue-target color cont ingencies. Experiment 2 puts this hypothesis to the test. Exper iment 2 Black and white schemat ic gaze cues were run in Exper iment 2 as matches for the black and white schemat ic arrow cues used in Exper iment 1. In all other aspects the design of Exper iment 2 replicated Experiment 1. Part icipants Forty additional naive participants were ass igned randomly and equally to two different groups: one group received black schemat ic eyes , one received white schemat ic eyes . Apparatus, Stimuli, Des ign & Procedure All parameters mirrored those in Experiment 1 with the following except ions that schemat ic black eyes and white eyes , served as fixation stimuli. These schemat ic eyes (measuring 2.6°) were constructed by combining a circle outline, with an inner filled-in circle representing the pupil (see Figure 6.1). C u e direction, target location, instructions, and number of trials were in keeping with Experiment 1, e.g., participants were correctly informed that cue direction and its color did not predict the location, color, or shape of the target. Resul ts Anticipations (RT< 100 ms), timed-out responses (RT> 1000ms), and incorrect key presses were classif ied as errors and were exc luded from the analys is . The errors occurred on only 4 % of trials and did not contradict the correct R T data. Correct mean R T s are summar ized in Figure 6.2B. In contrast to Exper iment 1, the spatial attention effect (cued R T < uncued RT) appears to be present for both cue-target color congruent and cue-target color incongruent stimuli. Separate within-subject 122 A N O V A s with cue validity (cued/uncued) and target color were performed for each group as in Exper iment 1. These ana lyses returned significant main effects of cue validity for each group [Black eyes F(1,19)=4.8,p<.05; White eyes F(1,19)=19.4,p< .0001]. However, unlike Exper iment 1, cue validity never interacted with target color [all Fs<2.2, all ps>.1]. Note that cue validity did not interact with target color even though a main effect of cue color was significant for both condit ions [Black eyes F(1,19)=14.0,p<01; White eyes F (1,19)=10.1,p<01]. No other effects were significant [all Fs<1]. A s suggested above, when group is included as a between subject factor it only interacts with target color, [F(1,38)=24.0,p<.0001], reflecting again that for each group R T is faster overall for the target color that is congruent with the color of the gaze cue. The only other significant effects are main effects of target color and cue validity [Fs>19,p<.0001; all other Fs<3.9,p>.05], the latter indicating that targets are responded to most quickly when they appear at the gazed-at location. D iscuss ion The main finding in Experiment 2 is that the reflexive spatial attention effect for gaze (cued R T < uncued RT) occurred for both the black eyes group and the white eyes group, and that this orienting effect for gaze direction was the same for congruent- and incongruently-colored targets. This stands in sharp contrast to the results of Experiment 1 where the spatial attention effect for both arrow groups occurred only for color congruent target stimuli. It is important to note that the overall cue-target color congruency effect observed in Experiment 1 reappeared in Exper iment 2. That is, the color congruency between gaze cues and target stimuli did affect overall RT , just as it did in Exper iment 1 for arrow cues . The critical difference is that in Exper iment 2 this factor did not interact with the 123 attention effect of gaze direction. Thus our observat ion that an interaction does not occur between cue validity and cue-target is not simply a matter of confirming the null. Both critical factors — reflexive orienting to gaze direction and cue-target color congruency — produced significant effects on performance. They did not, however, interact within or between groups. This positive finding supports the hypothesis that the attention effect for eyes is more strongly reflexive than the attention effect for arrows, and as such it is less vulnerable to the arbitrary cue-target color cont ingency attention effects that occur for arrows. Genera l D iscuss ion In recent years there has been an explos ion of interest in the finding that central directional cues can trigger reflexive shifts of attention to peripheral locations (e.g., (Fr iesen and Kingstone, 1998; Driver et al . , 1999; Langton & Bruce, 1999; Hommel et a l , 2001; Rist ic et a l , 2002; Fr iesen et a l , 2004; Rist ic & Kingstone, 2005). Originally, this work focused on the fact that reflexive orienting occurred to biologically relevant stimuli. This research usual ly concerned the effects of gaze direction (Fr iesen and Kingstone, 1998; Driver et al . , 1999; Langton & Bruce, 1999); but it a lso considered the attention effects of other biologically relevant stimuli like head direction and finger pointing (see Langton, Watt, and Bruce, 2000 for a review). More recently, research on the attention effects for directional cues has been extended to other directional cues such as arrows and words with spatial meaning (Tipples, 2002; Rist ic et al . 2002; Hommel et a l . 2001). Importantly, direct compar isons between these reflexive effects have been relatively infrequent, and when they have occurred they have tended to be grounded in the question of whether an effect that occurs for biologically relevant stimuli, like eyes , a lso occurs for other directional cues , like arrows (Baylis et a l . , 2005; T ipples, 2002; Rist ic et a l . 2002; Fr iesen et a l . 2004). 124 Based on these studies the conclus ion has been that the behavioral effects produced by gaze and arrow cues are very similar but the attention effect for gaze cues is more strongly reflexive than the attention effect elicited by arrow cues (Fr iesen et a l . , 2004). The present study took a different tack than those in the past. Rather than compar ing the effects of arrow cues against a reflexive attention effect that has been benchmarked by gaze , the present study compared the effects of eye gaze against a reflexive attention effect that has been benchmarked by arrows. Two exper iments examined the influence of cue-target color congruency on the reflexive attentional orienting effect that is observed when arrows and eyes are presented as spatially nonpredictive. The results of these exper iments represent a conceptual replication of Pratt and Hommel 's (2003) original finding that the emergence of a reflexive spatial attention effect for an irrelevant arrow cue is specif ic to a target stimulus that is colored the same as the arrow cue. Like Pratt and Hommel , the present findings demonstrate that the reflexive orienting effect of arrows is highly sensit ive to trial-by-trial changes in the attentional set for color that is establ ished by a color-irrelevant st imulus at fixation (see Pratt and Hommel , Exper iment 4). In addition, the present results demonstrate that the reflexive spatial attention effect for nonpredictive gaze cues appl ies equal ly to congruent and incongruently colored targets. This fundamental difference between the effects of arrows and eyes is consistent with the notion that the spatial attention effect for eyes is more strongly reflexive than the attention effect triggered by arrows. That is, the spatial attention effect triggered by gaze cues genera l izes across significant congruent and incongruent cue-target color cont ingency effects. 125 Note the conc lus ions that the reflexive orienting to gaze is uniquely resistant to a change in attentional set vis-a-vis the cue-target color congruency dovetai ls with the findings and conc lus ions of Fr iesen, Rist ic and Kingstone (2004). In that investigation, participants were informed that a target was likely to appear opposite to where the eyes looked. The outcome was that participants could not help but attend first to where the eyes were directed before shifting attention to the opposite location where a target was likely. In contrast, participants were able to avoid attending to where an arrow pointed, and simply shifted their attention to the likely opposite location. A similar finding of resistance to change in attentional set was also demonstrated by Rist ic and Kingstone (2005). In that study, participants were unable to avoid orienting to where an ambiguous stimulus was gazing once they had been informed that the stimulus depicted eyes. Thus we find that the present study joins a growing list of investigations that converge on the conclusion that the attentional effects of eyes are strongly reflexive, and that as such their effect on attentional orienting appears to be highly resistant to changes in attentional control settings. Conc lus ion There are at least three important implications of these data. First, they provide an important replication and extension of Pratt and Hommel (2003) who had first shown that trial-by-trial changes in cue-target attentional set can affect whether a reflexive orienting effect for arrows is observed. S e c o n d , the present data converge with the notion that gaze cues produce a more strongly reflexive effect than arrow cues (e.g., Fr iesen et al . , 2004). The reason that the attention effect of eyes is so powerful may stem from the fact that this attention effect is driven by the operation of brain mechan isms, like the S T S , that are dedicated to processing eye direction (e.g., Al l ison et al , 2000) and whose operation appears to be resistant to top-down modulat ion (e.g., 126 Vui l leumier et al , 2001; R e e s et a l , 2000). Finally, the current data agree with the proposal that there are a range of stimuli that can engage spatial attention reflexively, with the strength of this reflexive orienting effect varying ac ross stimuli (see also Rist ic & Kingstone, 2005). S o m e items, like eye direction, produce strongly reflexive effects and are thus highly resistant to modification by control settings. Other items, like arrows, and to a greater extreme numbers, are sensit ive to changes in the control setting that is adopted at any given time (e.g., Ristic, et al , in press; Gal fano, et a l , in press). Appreciat ion of this point may represent an important and positive step toward the development of a coherent theory of reflexive orienting and its impact on human performance. 127 Figure 6.1. Experiment 1: Arrow Black Arrow White Arrow > / Experiment 2: Eyes Black Eyes White E y e s cue 105ms SOA 212ms target 120ms Figure 6.1. Illustration of stimuli and sample sequence of events. In Experiment 1, a black or a white arrow served as a fixation stimulus. In Experiment 2, a pair of black or white schematic eyes served as fixation stimulus. Note that the stimuli are not drawn to scale. 128 Figure 6.2. Figure 6.2A. Experiment 1 Results. Black Arrow White Arrow 460 450 440 430 h or § 4 2 0 CD 410 h 400 h 390 • white target • black target 460 450 | 440 H 430 or S 4 2 0 410 Cued Uncued C u e Validity Figure 6.2B. Experiment 2 Results. Black Schematic Eyes Cued Uncued C u e Validity White Schematic Eyes 400 h • white target • black target 390 Cued Uncued Cued Uncued C u e Validity C u e Validity Figure 6.2. Experiment 1 and Experiment 2 cued and uncued RTs as a function of target color. Figure 6.2A illustrates performance for Black and White arrow groups. Figure 6.2B shows Black and White gaze groups. 129 References Al l ison, T. Puce , A , & McCar thy G . (2000).Social perception from visual cues : role of S T S region. Trends in Cognit ive Sc iences , 4, 267-278. Batki , A . , Ba ron -Cohen , S . , Wheelwright, S . , Connel lan , J . , & Ahluwal ia, J . (2000). Is there an innate modu le? Ev idence from human neonates. Infant Behavior & Development. 23 , 2 2 3 - 2 2 9 . Bay l iss , A . P . , di Pel legrino, G . & Tipper, S . P. (2005). S e x differences in eye-gaze and symbol ic cueing of attention. Quarterly Journal of Exper imental Psycho logy , 58A, 631-650. Driver, J . , Davis, G . , Ricciardel l i , P. , Kidd, P. , Maxwel l , E., & Baron -Cohen , S . (1999). G a z e perception triggers visuospat ial orienting by adults in a reflexive manner. V isual Cogni t ion, 6, 509-540. Farroni, T., Cs ib ra , G . , S imion, F., & Johnson , M. H. (2002). Eye contact detection in human from birth. Proceed ings of the National A c a d e m y of Sc i ences , 99, 9603-9605. F ischer, M.H. , Cas te l , A . D . , Dodd, M.D., & Pratt, J . (2003). Perceiv ing numbers causes spatial shifts of attention. Nature Neurosc ience, 6, 555-556. Fr iesen, C . K., & Kingstone, A . (1998). The eyes have it!: Ref lexive orienting is triggered by nonpredictive gaze . Psychonomic Bulletin & Review, 5, 490-495. Fr iesen, C . K., Ristic, J . , & Kingstone, A . (2004). Attentional effects of counterpredict ive gaze and arrow cues . Journal of Experimental Psycho logy : Human Percept ion & Per formance. 30, 319-329. Gal fano, G . , Ruscon i , E., & Umilta, C . (in press). Number magnitude orients attention, but not against one 's will. Psychonomic Bulletin & Rev iew. Gibson , B. S . & Kingstone, A . (in press). V isua l attention and the semant ics of space : Beyond central and peripheral cues . Psycholog ica l Sc ience . Hommel , B., Pratt, J . , Co lza to , L. & Godi jn, R. (2001). Symbol ic control of visual attention. Psycho log ica l Sc ience ,12 . 360-365. Hood, B. M., Wi l len, J . D., & Driver, J . (1998). Adult 's eye trigger shifts of visual attention in human infants. Psycholog ica l Sc ience , 9, 131-134. Kobayash i H, & Kosh ima S . (1997). Unique morphology of the human eye. Nature. 387, 767-768. Langton, S . R. H., & Bruce, V . (1999). Reflexive socia l orienting. V isua l Cogni t ion. 6, 541-567. Langton, S . R. H., Watt, R. J . , & Bruce, V . (2000). Do the eyes have it? C u e s to the direction of socia l attention. Trends in Cognit ive Sc iences . 4, 50-59. Pratt, J . & Hommel , B. (2003). Symbol ic control of v isual attention: The role of working memory and attentional control settings. Journal of Exper imental Psycho logy : Human Percept ion & Per formance, 29, 835-845. R e e s , G . , Wojciulik, E, C larke, K, Husain, M. , Frith C . & Driver, J . (2000). Unconsc ious activation of visual cortex in the damaged r ight-hemisphere of a parietal patient with extinction. Brain. 123,1624-1633. Rist ic, J . , Fr iesen, C . K., & Kingstone, A . (2002). Are eyes spec ia l? It depends on how you look at it. Psvchonomic Bulletin & Rev iew, 9, 507-513. Rist ic, J . , & Kingstone, A . (2005). Taking control of reflexive socia l attention. Cogni t ion, 94, B 5 5 - B 6 5 . Ristic, J . , Wright, A . & Kingstone, A . (in press). The number line effect reflects top-down control. Psvchonomic Bulletin & Rev iew. 131 Tipples, J . (2002). Eye gaze is not unique: Automat ic orienting in response to noninformative arrows. Psychonomic Bulletin & Rev iew, 9, 314-318. Vui l leumier, P. et al (2001). Neural fate of seen and unseen faces in visuospat ial neglect: A combined event-related functional MRI and event-related potential study. Proceed ings of the National A c a d e m y of Sc ience , U S A , 98, 3495-3500. Watanabe, K. (2002). Ref lexive attentional shift caused by indexical pointing gesture. V is ion S c i e n c e s Society, Saraso ta , Flor ida, U S A , May 2002. 132 S E C T I O N III: I M P L I C A T I O N S F O R T R A D I T I O N A L M E A S U R E S O F A T T E N T I O N Introduction 133 E a c h study presented in this dissertation has demonstrated that a directional, spatially nonpredictive attentional cue presented at central fixation will trigger a reliable reflexive shift in spatial attention toward a peripheral location (Ristic & Kingstone, 2005; Rist ic et al , 2005; Rist ic & Kingstone, 2002; Fr iesen, Rist ic & Kingstone, 2004; Rist ic, Wright & Kingstone, Chapter 6). This finding contrasts with the traditional conceptual izat ion and measurement of attentional orienting. In the standard cuing paradigm, reflexive spatial orienting is triggered by presenting a spatially nonpredictive st imulus in the periphery, and volitional orienting is engaged by presenting a spatially predictive directional cue at central fixation (e.g., Posner , 1980; Jon ides , 1981). Importantly, the attention effect of a central spatially nonpredict ive central st imulus, such as a gaze cue or arrow cue, displays properties that are suggest ive of both reflexive orienting and volitional orienting. In keeping with reflexive orienting these cues are spatially nonpredict ive and their effects occur shortly after a cue is presented. In keeping with volitional orienting these cues produce a susta ined performance benefit at the cued location that does not appear to be replaced by the inhibition of return (IOR) effect. In this final sect ion of the thesis, Chapter 7, the reflexive attention effect when it is triggered by a central nonpredict ive cue is compared directly against traditional measures of reflexive and volitional orienting, that is, orienting to nonpredict ive peripheral onsets and predictive central cues , respectively. To achieve this end, I developed a novel paradigm that al lows an unbiased compar ison between two different types of attentional cues by integrating them simultaneously with the s a m e task. The result is that on any given trial, two attention cues are presented, and they may either diverge spatially (cuing different locations) or converge spatially (cueing the same location). Note that this manipulation al lows for an estimation of the attentional effects 134 when each cue indicates a different spatial location and the attentional effects when the two cues indicate the s a m e spatial location. Part icipants are a lways asked to detect a single target that occurs in one of four possible target locations following different st imulus onset asynchrony ( S O A ) intervals (100, 300, 600 or 900ms) . In this way one can purchase a clear picture of the dynamics of the two orienting p rocesses , and how they may impact each other, across time. The reasoning is straight-forward. If the attentional effects elicited by the two cues are independent, then the attention effect of each cue will proceed in a manner that is unaffected by the location that is cued by the other st imulus or its spatial predict iveness. O n the other hand, if the cues are not independent, the performance of one cue will be impacted by the location or spatial predict iveneess of the other cue. The results of this investigation demonstrate independence between each of the cues (nonpredictive central, nonpredictive peripheral, and predictive central cues) . One of the main implications of this finding is that indicates that reflexive orienting to a nonpredictive central cue is distinct from reflexive orienting to an abrupt peripheral onset or volitional orienting to a predictive central cue. The broader implications of this study, and those that have preceded it in Chapters 2-6, are considered in depth in the Genera l D iscuss ion that forms Chapter 8. 135 References Fr iesen, C . K., Rist ic, J . , & Kingstone, A . (2004). Attentional effects of counterpredictive gaze and arrow cues . Journal of Experimental Psycho logy: Human Percept ion & Per formance, 30, 319-329. Jon ides , J . (1981). Voluntary versus automatic control over the mind's eye's movement. In J . B. Long and A . D. Badde ley (Eds.) , Attention & Per formance IX (pp. 187-203). Hi l lsdale, N J : Er lbaum. Posner , M. I. (1980). Orienting of attention. Quarterly Journal of Exper imental Psycho logy. 32, 3-25. Rist ic, J . , Fr iesen, C . K., & Kingstone, A . (2002). Are eyes spec ia l? It depends on how you look at it. Psychonomic Bulletin & Review, 9, 507-513. Rist ic, J . , & Kingstone, A . (2005). Taking control of reflexive socia l attention. Cogni t ion, 94, B55 -B65 . Rist ic, J . , Mottron, L., Fr iesen, C . K., larocci, G , Burack, A . J . , & Kingstone, A . 2005). E y e s are specia l but not for everyone: The case of aut ism. Cognit ive Brain Resea rch , 24, 715-718. 136 C H A P T E R 7 Nonpredict ive Central C u e s Trigger Independent Ref lexive Effects A version of this chapter will be submitted for publication. Rist ic, J . & Kingstone, A . Nonpredict ive central cues trigger independent reflexive effects. 137 The allocation of spatial attention is often conceptual ized as being committed reflexively or volitionally. Ref lexive orienting is understood to occur exogenously , that is, in response to an external stimulus event; and volitional orienting is understood to occur endogenously , in response to one's internal goals and expectat ions (Posner , 1978; Klein, Kingstone and Pontefract, 1992). This division of attentional select ion has been formal ized by two prominent behavioral testing procedures — the peripheral cuing task and the central cuing task. These two tasks set out to isolate, engage and measure reflexive and volitional orienting by manipulating both the spatial position of the attentional cue and its ability to predict where a response target is likely to appear. In the peripheral cuing task, the abrupt onset of a spatially nonpredict ive cue at a parafoveal location is understood to engage reflexive spatial orienting. This type of orienting is character ized by targets being responded to more quickly at the cued versus noncued location when the cue-target stimulus onset asynchrony ( S O A ) is less than about 300 ms .When the S O A is longer, R T to targets tend to be shorter at the uncued versus cued location. This latter effect is known as "inhibition of return", or IOR, and is thought to reflect the fact that attention is withdrawn from the cued location and then is inhibited from returning there (Posner , 1980; Posne r & C o h e n , 1984). In the central cuing task, a spatially predictive cue presented at a foveal location is understood to engage volitional spatial orienting. This effect is measured by typically presenting a central arrow cue that predicts the location of a response target. This type of orienting is character ized by targets being responded to more quickly at the location likely to contain the target (Jonides, 1981). S ince the development of the expectancy is a cognitively demanding task, the effect of volitional attention are typically seen first when the cue-target S O A approximates 300 ms (Muller & Rabbitt, 1989). Voluntary orienting is never character ized by the presence of an IOR effect as participants have 138 an incentive to maintain attention at the location that is likely to contain the target (Taylor & Klein, 1998). Severa l recent studies have demonstrated a variety of central cues , such as eyes looking to the left or right, or an arrow pointing to the left or right, will produce shorter R T latencies for targets at the cued versus noncued location (e.g., Eimer, 1997; Tipples, 2002; Rist ic et a l , 2002). B e c a u s e the directional cue does not predict the target location and the attention effect emerges shortly at cue-target S O A s as short as 100 ms, this cue is understood to engage reflexive orienting. Unlike the orienting effect for peripheral cues , however, this facilitory R T effect persists well beyond 500 ms and is not accompan ied by the emergence of an IOR effect (e.g., Tipples, 2002; Fr iesen & Kingstone, 2003). In this important way, the reflexive orienting effect a lso shares properties of volitional orienting as it is normally def ined, i.e., orienting is engaged by a central cue, its facilitory effect is long-lasting, and IOR does not occur. In light of this overlap between reflexive orienting to central cues and the more traditional understandings of reflexive orienting to peripheral cues and volitional orienting to central cues , the present study set out to examine the relationship between these three attention tasks and the orienting effects they are thought to engage. This effort brings forward two main quest ions. First, what is the relationship between reflexive orienting when it is triggered by a peripheral cue and when it is triggered by a central c u e ? In particular, should reflexive spatial orienting to peripheral and central cues be thought of as different ways of engaging the s a m e underlying attention sys tem, or two different attention sys tems? S e c o n d , what is the relationship between volitional and reflexive orienting when they are both engaged by a central cue? The literature suggests that volitional and reflexive attention, as elicited by the two c lass ic tasks are subserved by two different brain networks (e.g., Corbetta & Shu lman , 2002). However, 139 the link between volitional orienting and reflexive orienting when both cues are shown at central fixation is an area that has not been investigated. The present study addressed these quest ions by taking the three cuing tasks and pitting one against the other in three separate test condit ions: Nonpredict ive Peripheral - Nonpredict ive Arrow ( N P - N A ) cues ; Predict ive Central - Nonpredict ive Arrow ( P C - N A ) cues ; Nonpredict ive Peripheral - Predict ive Central ( N P - P C ) cues . This approach is grounded in additive factors logic, which provides a wel l -establ ished method for understanding the relationship between cognitive p rocesses (Sternberg, 1969; Posner 1978). Effectively, if two cognit ive p rocesses , such as two types of reflexive attention or reflexive and volitional attention, are mediated by different mechan isms, their effects should co-occur without interacting, i.e., they are independent. In contrast, if the two processes are mediated by a common mechan ism, their effects will combine in an interactive manner, e.g., they may interfere or accentuate one another indicating that they are not independent. Note that any suggest ions of independence for a particular type of cuing can also be tested by examining whether the orienting effect var ies as a function of the cue condit ion it was paired with. S o , for example, if the effects of nonpredict ive peripheral onsets do not interact with the effects of nonpredict ive arrows or predictive central cues (thereby suggest ing independence of peripheral orienting from the other two forms of orienting), it follows that the orienting effect for nonpredictive onsets should remain the s a m e when paired with the nonpredict ive arrow cue and the predictive central cue. In order to gain a detailed insight into performance across time, R T was measured at four different cue-targets S O A s (100, 300, 600, and 900 ms). This al lows one to a s s e s s whether R T facilitation occurs early at a cued location as expected for reflexive orienting, and also to examine whether the effects diverge at later S O A s , e.g., 140 an IOR effect at the cued location as expected for nonpredictive peripheral (NP) cues versus a facilitory effect at the cued location as expected for predictive central (PC) cues . Finally it should be noted that because nonpredictive arrow (NA) cues produce reflexive orienting when they are presented centrally, arrows are not suitable central cues for the P C cue condit ion. It has now been demonstrated in two studies that number cues presented at central fixation do not produce reflexive orienting when they are made spatially nonpredictive, but that they readily enable volitional orienting when they are made spatially predictive (Ristic & Kingstone, in press; Rist ic, Wright & Kingstone, in press). Hence in the present study spatially predictive number cues were used in the P C condit ion. Method Part icipants The intention had been to ass ign 15 participants to each N P - N A and N P - P C group and 30 participants to P C - N A group, to ensure adequate sampl ing in the condit ion, which presents two cues foveally. Due to participant over schedul ing 19 participants were ass igned to the N P - N A group, 15 to the N P - P C group and 30 to the P C - N A group. Al l participants were naive to the purpose of the experiment. Apparatus and Stimuli Test ing was performed on a Macin tosh power P C computer connected to a 15-in monitor set to black and white. All stimuli were black line drawings shown on a white background. Per ipheral cues were created by thickening the outline of one of the 2° x 2° placeholder boxes, posit ioned 7.5° away from central fixation along horizontal and vertical p lanes. A central arrow cue was created by combining a straight line (2°) with an arrowhead and an arrowtail. The central number cue (either 1, 3, 6, or 9) measured 2° in 141 height and 1.2° (3, 6, 9) or 0.3° (1) in width. The target was an asterisk, subtending 0.8° of visual angle, appear ing with 7.5° eccentricity, as measured from the center fixation to the center of the target. The stimuli, sample sequence of events and cue validity condit ions for the N P - N A condit ion are illustrated in Figure 7.1, for the P C - N A condition in Figure 7.2 and for the N P - P C condit ion in Figure 7.3. Des ign E a c h condition was as a within-subjects design, setup as a four location cuing task. O n each trial, participants were presented with two attentional cues simultaneously. In the N P - N A condit ion, a spatially nonpredict ive peripheral onset cue and a spatially nonpredictive central arrow cue were shown. In the P C - N A condit ion, a spatially predictive central number cue and a spatially nonpredict ive central arrow cue were shown. Finally, in the N P - P C condit ion, a spatially nonpredict ive peripheral onset cue and a spatially predictive central number cue were shown. A single target appeared at one of the four target locations following one of four randomly determined S O A intervals of 100, 300, 600 or 900 ms. In all condit ions peripheral onset cues were presented for 90 ms while central cues (both arrow and number) were presented for the duration of the trial. In the N P - N A condit ion, the position of the peripheral cue and the direction of the arrow cue was determined randomly, with the target appear ing with equal probability at each location (p=.25). In the P C - N A and N P - P C condit ions, the central number cue indicated the correct target location with .77 probability. Number 1 predicted a target occurring on the top; number 6 a target on the bottom, number 3 a target on the right and number 9 a target on the left. In both of these condit ions, the peripheral onset cue and the central arrow cue indicated a correct target location equal ly often with .07 probability. It is worth noting that s ince P C - N A condition utilized two foveal cues , the 142 number cue was a lways posit ioned above fixation and the arrow cue was a lways posit ioned below fixation at an equal d istance of 3.5°, as measured from the center of the number cue to the center of an arrow cue. O n each trial, two attentional cues either diverged spatially, indicating two different spatial locations, or converged spatially, indicting the s a m e spatial location. A s illustrated in F igures 7.1, 7.2, and 7.3 in the spatially divergent condit ion, the effect of each cue was a s s e s s e d by a compar ison against uncued trials in which the target appeared at one of the two locations that were not cued. In the spatially convergent condition the effect of both cues was compared against uncued trials in which the target appeared at one of the three locations that were not cued. Procedure E a c h trial began with a presentation of a fixation display for 1000 ms. Then , the two attentional cues appeared. In the N P - N A condit ion, one of the four boxes was cued for 90 ms and an arrow, pointing in one of the four directions appeared at the center. In the P C - N A condit ion, an arrow cue, pointing in one of the four directions and a number cue (1, 3, 6, or 9) appeared at the center. In the N P - P C condit ion, one of the four boxes was cued for 90 ms and a number cue (1, 3, 6 or 9) appeared at the center. The onset of the two cues was a lways s imul taneous. Fol lowing a randomly selected S O A of 100, 300, 600 or 900 ms, a target demanding a simple detection response appeared at one target location (left, right, up, down). The trial was terminated on response or after 2600ms had e lapsed , whichever came first. The intertrial interval was 675ms. R T was measured from the onset of the target until the response key was p ressed . Randomly , on approximately 6% of the trials, a target was not presented (catch trial) and participants were required to withhold a keypress detection response. 143 Part icipants were seated centered with respect to the computer sc reen at an approximate distance of 57 cm . They were instructed to maintain central fixation throughout the experiment and to press the spacebar with the index finger of their preferred hand as fast and as accurately as possible as soon as they detected the target. Al l participants were informed, and it was confirmed that they understood, the predict iveness of each cue. C u e direction, target position, and S O A were presented equally and in random order. All participants completed a total of 960 trials divided into 16 testing blocks of 60 trials in each condit ion. Ten practice trials were run at the beginning. Resul ts One participant was exc luded because of extremely deviant R T performance. The mean correct R T for each cue condition as a function of cue validity and S O A were computed. Anticipations (RT<100), timed-out responses (RT>1000), incorrect key presses, and false alarms were infrequent and were removed from the analys is . The error rate on target present trials was 1.05%, 0.6% and 1.95% with false alarm rate on target absent trials of 1.64%, 2 % and 4 % for the N P - N A , P C - N A and N P - P C condit ions respectively. The behavioral effects for the nonpredictive peripheral cuing task, nonpredictive arrow cuing task and predictive central cuing task are well establ ished and have been replicated on numerous occas ions . A s outlined in the introduction, nonpredict ive peripheral onset cues elicit reflexive orienting that is marked by a b iphasic R T pattern with R T facilitation for short cue-target intervals which is replaced by IOR at longer cue-target delays (e.g., Posner , 1980; Posne r & C o h e n , 1984). Nonpredict ive central cues elicit reflexive orienting with a R T pattern marked by facilitation occurring for early and late cue-target intervals which is never replaced by IOR (e.g., Tipples, 2002; Rist ic et al , 144 2002). Finally, predictive central cues produce volitional orienting that is marked by facilitation for predicted targets for de lays exceeding 300ms that is never replaced by IOR (e.g., Jon ides , 1981; Taylor & Kle in, 1998). First we examined for the presence of each of these well-replicated effects and their stability as a function of the cue they were paired with. Figure 7.4 plots the mean R T s for each individual cue type as function of its cue pairing, cue validity and S O A . A s suggested by Figure 7.4, the orienting effect for each cue type is what one would expect to find based on past research and, most importantly, the effects appear to be stable for each cue type regardless of the cue it is paired with. This observat ion was confirmed by a mixed analys is of var iance ( A N O V A ) conducted on each cue condit ion. The first analys is compared the effects elicited by a nonpredictive peripheral onset cue across the nonpredictive arrow and predictive central cue condit ions, with type of cue pairing (NA, P C ) as a between subject factor and cue validity (cued; uncued), and S O A (100, 300, 600, 900) as within-subject factors. A second A N O V A similarly examined the attentional effects elicited by a nonpredictive arrow across the nonpredict ive peripheral and predictive number condit ions (NP, P C ) . A third and final A N O V A in a similar manner compared the effects elicited by a predictive central cue across the nonpredict ive peripheral onset and nonpredictive arrow condit ions (NP, NA) . The first A N O V A confirmed that nonpredictive peripheral onset cues elicited reflexive orienting effects marked by early facilitation and later inhibition, regardless of whether they were paired with a nonpredictive arrow or a predictive central cue. That is, there was a significant S O A x cue validity interaction [F (3, 96)=8.5, p<.0001] which did not interact with cue type [F (3, 96)=1.3, p>.25]. Four two-tailed paired t-tests verified that the facilitation effect at the cued location was significant at 100 ms (t(33)=-2, p=.05), d isappeared at 300 ms (t(33)= -1.4, p>1.7), and gave way to the IOR effect at 145 the two longest S O A s (both ts>3, ps<.01). The only other significant effect w a s S O A [F (3, 96)= 19.6, p<.000], reflecting a standard foreperiod effect whereby overal l R T decl ines as S O A increases (Bertelson, 1967). No other effects or interactions were reliable (all Fs<1.6, all ps>.2). In sum, these data indicate that nonpredict ive peripheral onset cues produced a standard b iphasic R T pattern of reflexive orienting that was stable across the two cue condit ions (NA, P C ) . The second A N O V A confirmed that a nonpredictive arrow cue elicited the predicted reflexive orienting effect, marked by R T facilitation across all S O A s . This effect occurred regardless of whether the nonpredictive arrow was paired with a nonpredictive peripheral onset cue or a predictive central cue. Main effects of S O A [F (3, 138)=40, p<0001] and cue validity [F (1, 46)=12, p< .01] were reliable, with no other significant effects or interactions emerged, including those involving the cue-type pairing (all Fs<1). The final A N O V A confirmed that a predictive central cue elicited volitional orienting, marked by facilitation for targets appear ing at the cued (likely) location, that emerged at approximately 300 ms S O A and grew and persisted thereafter, regardless of whether the predictive cue was paired with the nonpredictive peripheral onset or the nonpredictive arrow cue. There were the standard main effects of S O A [F (3, 129)=65, p<0001] and cue validity [F (1, 43)=24, p< .0001] as well as an S O A x cue validity interaction [F (3, 129)=5.8, p< .001] reflecting the development of the cuing effect over time. Paired two-tailed t-tests verified that the facilitation effect at the cued location was nonsignif icant at 100 ms (t(44)=-.8, p>.4), and significant thereafter (all ts>-2.7, ps<.01). No other effects or interactions were significant, including any involving the cue-type pairing [all Fs<1.7, all ps>.2]. In sum, these data indicate that predictive central number 146 cues engaged a c lass ic volitional orienting effect that was stable across cue-type pairings. In all the above ana lyses the expected effects for each cue type was observed across the S O A condit ions, and these effects did not vary as a function of the cue-type that they were paired with. Col lect ively, these ana lyses strongly suggest that the attention effects for each cue type are independent from each other. That is, the nonpredict ive peripheral cues and nonpredictive arrow cues trigger reflexive orienting effects that are independent from one another as well independent from the volitional orienting effects engaged by predictive central cues . To verify this interpretation, three within-subjects A N O V A s were conducted for each cue condition ( N P - N A , P C - N A , and N P - P C ) compar ing the effects of each individual cue when they indicated different spatial posit ions and when they indicated the same spatial condit ion. If their effects are independent, then the sum of their attention effects (uncued R T - cued RT) when divergent locations are cued should equal the magnitude of the attention effect (uncued R T - cued RT) when they converge on the same location. To anticipate the outcome of these ana lyses , the attention effects for all cue-types were found to operate independently. These data are illustrated in Figures 7.5-7.7. Nonpredict ive Per ipheral - Nonpredict ive Arrow (NP-NA) C u e s . A within-subjects A N O V A compar ing peripheral cued and uncued R T s in the spatially divergent case revealed a significant main effect of S O A [3, 54)=24.5, p<.0001] and an interaction between cue validity and S O A [F(3, 54)=3.6, p<.05], again reflecting the emergence of an IOR effect. 147 A separate A N O V A compar ing the arrow cued and unced R T s revealed significant main effects of S O A [F(3, 54)= 30, p<0001] and cue validity [F(1, 18)= 19, p< .001] and no interactions (p>.2). A n A N O V A with cue validity (both cued ; both uncued) and S O A conducted on the spatially convergent condition revealed a significant main effect of S O A [F(3, 54)=7, p<.001] and S O A x cue validity interaction [F(3, 54)=5, p<.05] indicating that the early facilitation effect is later abol ished, presumably because the facilitation effect for a nonpredictive arrow is countered by the addition of an IOR effect triggered by a nonpredictive peripheral cue. To test the notion that the sum of the cuing effects for divergent N P and N A cues approximates the cuing effect when the cues are convergent, the magnitude of the attention effect (uncued R T - cued RT) for the divergent cues was calculated and summed for each S O A . These data were compared against the attention effect for each convergent cue at each S O A . This resulted in a 2 (divergent sum/convergent) x 4 ( S O A ) within-subject A N O V A that produced only a significant main effect of S O A [(3, 54)=7.4, p< .001] reflecting that the attention effect was present at the short S O A s but that it was abol ished later by the IOR effect. Al l other Fs<1. Thus , the attention effects of the divergent cues add together and equal the magnitude of the attention effects for the convergent cues for all S O A s . The equality of the overall magnitudes between divergent and convergent cues is illustrated in Figure 7.5B. Nonpredict ive Arrow - Predict ive Central ( P C - N A ) C u e s . A within-subjects A N O V A compar ing arrow cued and uncued R T s in the spatially divergent case revealed a significant main effect of S O A [F (3, 84)=19, p<.0001] and an effect of cue validity that brushed signif icance [F(1, 28)= 3.5, p<.06] due to the inclusion of one outlier. W h e n this participant is exc luded the nonpredict ive arrow effect becomes highly significant, with a 148 p-value of 0.01. This fact, coupled with the previous observat ion that this arrow effect is equivalent to the statistically significant arrow effect in the N P - N A condit ion, suggests that the N P arrow effect here is reliable. A separate A N O V A compar ing cued and uncued R T s for predictive number cues revealed significant main effects of S O A [F (3, 84)= 38.5, p<.0001], cue validity [F (1, 28)=12.3, p<01] and an S O A x cue validity interaction [F (3, 84)=4.2, p<01] . A n A N O V A conducted on the spatially convergent case revealed significant effects of S O A [F(3, 84)=33.3, p<0001] and cue validity [F (1, 28)=31, p<05] with no interaction (p>.1). The magnitudes of divergent and convergent attention effects were compared in a 2 (divergent sum/convergent) x 4 ( S O A ) within-subject A N O V A as before. This analys is returned no significant effects or interactions (all Fs<1.7, all ps>.1) indicating that the sum of the attention effects in the divergent cue c a s e are equivalent to the attention effects observed in the convergent cue case . The equality of the overall magnitudes between divergent and convergent cues is shown in Figure 7.6B. Nonpredict ive Peripheral - Predict ive Central ( N P - P C ) C u e s . A within-subjects A N O V A compar ing peripheral cued and uncued R T s in the spatially divergent case revealed a significant main effect of S O A [F (3, 42)=5.4, p<.01] and an interaction between cue validity and S O A [F(3, 42)=4.2, p<.01] again demonstrat ing the emergence of the IOR effect. A separate A N O V A compar ing cued and uncued R T s for predictive number cues revealed significant main effects of S O A [F(3, 42)= 28, p< 0001] and cue validity [F (1, 14)=10, p< .01] with no interaction (p>.05). A n A N O V A conducted on the spatially convergent case revealed significant effects of S O A [F(3, 54)=7, p<.001], validity [F (1, 14)=7.4, p<05] as well as an S O A x cue validity interaction [F (3, 42)=3.8, p<05]. 149 The magnitudes of divergent and convergent attention effects were compared in a 2 (divergent sum/convergent) x 4 ( S O A ) within-subject A N O V A as descr ibed above. This analysis returned no significant effects or interactions (all Fs<1.5, all ps>.2) indicating that the sum of the attention effects in the divergent cue case are equivalent to the attention effects observed in the convergent cue case . The equality of the overall magnitudes between divergent and convergent cues is shown in Figure 7.7B. Discuss ion This study set out to answer two main quest ions. What is the relationship between reflexive orienting when it is triggered by a peripheral cue and when it is triggered by a central cue? A n d what is the relationship between volitional and reflexive orienting when they are both engaged by a central cue? W e also examined the c lass ic relationship between reflexive orienting to a peripheral cue and volitional orienting to a predictive central cue. The results of this investigation are clear-cut. Al l results show consistently that each of the three cuing effects — nonpredictive peripheral (NP) , nonpredict ive arrow (NA) and predictive central (PC) — produce the R T effects that would be expected if the cues were presented in isolation. The new finding is that these attention effects can co-occur in an independent manner. This was demonstrated both in the stability of the effects across cue-type pairings and within cue-type pairings. Ac ross cue-type pairings it was found that peripheral onset cues triggered the same c lass ic b iphasic R T effect regardless of whether peripheral cues were paired with a nonpredictive arrow cue that engages reflexive orienting or a predictive central cue that engages volitional orienting. Similarly, nonpredictive arrows triggered the expected early and prolonged R T facilitation effect at a cued location that was evident regardless of whether the nonpredict ive arrow w a s paired with a nonpredictive peripheral cue that 150 triggers a biphasic R T effect or a predictive central cue that engages volit ional orienting. Finally, a predictive central cue engaged volitional orienting that was the s a m e regardless of whether it co-occurred with reflexive orienting to a peripheral onset cue or volitional orienting to a central cue. This independence of the attention effects across different groups of cue-pair ings was cross val idated by a within-subjects compar ison of the magnitude of the attention effects when the two cues diverged spatially and when they converged spatial ly on the same location. For each cue pairing the results showed consistently that the attention effects for divergent cues summed to equal the magnitude of the attention effects when the cues converged on the s a m e location. Together then these data strongly suggest that spatially nonpredict ive cues elicit reflexive orienting effects that are independent of each other, and that each of these reflexive attention effects is independent from the volitional orienting effect that is generated endogenous ly in response to a spatially predictive symbol ic central cue. Whi le there is good ev idence from past studies that reflexive orienting to a nonpredictive peripheral cue and volitional orienting to a central cue are independent and subserved by qualitatively different brain mechan isms (e.g., Corbetta & Shu lman, 2002), this is the first study to demonstrate that reflexive orienting to nonpredictive peripheral and central cues occur independently, and that reflexive orienting to nonpredict ive central cues is distinct from volitional orienting to central cues . These novel f indings offer a resolution to a potentially contentious issue in the literature regarding the nature of the attentional effects elicited by spatially nonpredict ive central cues . Our data show that these effects are independent from both reflexive and volitional effects elicited by two c lass ic tasks and as such argue against the hypotheses that the attention effects elicited by reflexive central cues represent a hybrid form of 151 c lass ic reflexive and volitional orienting (Klein & Shore , 2000; Klein, 2004) or even more mundanely a variant of c lass ic volitional orienting (Vecera & R izzo , 2006). The first hypothesis predicts interactive effects between nonpredictive arrow cues and both nonpredictive peripheral cues and predictive central cues . This was clearly not the case . The second hypothesis predicts that the effects of nonpredictive arrow cues will interact with the effects of predictive central cues . Aga in , this prediction was disconf i rmed. Instead, our data support the conclus ion that central nonpredict ive arrows elicit reflexive attention effects that are independent from the reflexive effects elicited by peripheral cues and the volitional orienting effects elicited by predictive central cues . These data carry a number of important implications for the field's character izat ions of reflexive and volitional orienting. First, they suggest that reflexive orienting in a cuing paradigm need not be character ized by a behavioral response that occurs in response to a peripheral cue, nor can volitional orienting be character ized simply as a behavioral response that occurs in response to a central cue. This fol lows from the result that central nonpredict ive cues produce reflexive orienting from a foveal location. S e c o n d , the data suggest that reflexive orienting need not be character ized as a behavioral response that is fol lowed by IOR and volitional orienting cannot be character ized simply as a behavioral response that is never fol lowed by IOR, i.e., by a sustained facilitation at a cued location. This follows from the result that central nonpredict ive cues do not a trigger IOR and exhibit sustained facilitation at the cued location. Finally, these points above suggest that an attribution of reflexive orienting to ventral brain sys tems that are activated by peripheral onsets and superior sys tems that are engaged by symbol ic central cues may need to be reexamined (e.g., Corbetta & Shu lman, 2002). 152 The present data a lso open interesting quest ions for future investigation. In light of the behavioural d issociat ions between reflexive and volitional orienting to central and peripheral cues demonstrated here, one wonders what brain mechan isms are critical to the independence of the attention effects (and conversely, what brain mechan isms are shared between cues and are therefore not critical to behavioural d issociat ions). Currently there are no investigations that have systematical ly examined the reflexive and volitional orienting effects that occur in response to spatially nonpredict ive central and peripheral cues or spatially nonpredictive and predictive central cues . There is a lso the outstanding question concerning the relationship between reflexive and volitional orienting when eye movements are withheld and when they are executed. The present investigation examined covert attentional orienting controlling for eye movements by using a task that requires detection of suprathreshold targets (e.g., Posner , 1980; Kingstone & Pratt, 1999; Fr iesen et al , 2004;). Typical ly, however, people move their eyes toward those things that are of interest (Findlay & Gilchrist, 2003). The extent that the present f indings apply to this performance domain, and by extension, to real-world behavior is unknown. Finally, it is worth noting that while a range of central attentional cues (e.g., eye direction, arrow direction, head direction, finger pointing, words with spatial meaning) produce similar attention effects in simple behavioral tasks, it appears that different c lasses of central cues produce reflexive attention effects that range from strongly reflexive (resistant to cognitive control) to weakly reflexive (easily modified by top-down set). For example, the attentional effects of central eye direction are resistant to interruption (Fr iesen et al , 2004), contextual top-down modulation (Ristic & Kingstone, 2005), and attentional control settings (Ristic, Wright & Kingstone, submitted [see also Chapter 7]). In contrast, attentional effects of central arrow cues are not resistant to 153 interruption (Fr iesen et a l , 2004) and are highly amenab le to changes in attentional control sett ings. This ra ises the very real possibility that a range of central cues engage reflexive attention differently by engaging dissimilar brain mechan isms. 154 Figure 7.1. Fixation Display 1000ms Spatial ly Divergent Condit ion Spatial ly Convergent Condit ion Peripheral Cue(90ms) Arrow Cue(until response) SOA 100, 300, 600, or 900ms Target Onset 2600ms onset cued arrow cued uncued both cued both uncued Figure 7.1. Illustration of stimuli (not to scale) and sample sequence of events for the nonpredictive peripheral - nonpredictive arrow (NP-NA) cues condition. Every trial began with a 1000 ms presentation of a fixation display followed by a simultaneous presentation of two cues: a spatially nonpredictive peripheral onset cue and a spatially nonpredictive central arrow cue. Both the peripheral and the central arrow cues were spatially nonpredictive with a target appearing at the cued location on only 25% of the trials. On any given trial, a peripheral onset and arrow cue could diverge spatially (Spatially Divergent) or converge spatially (Spatially Convergent). The peripheral cue was presented for 90 ms while the central arrow cue remained on the screen for the duration of the trial. There were three cue validity conditions in the Spatially Divergent case: onset cued trial, where the target appeared at the location indicated by the peripheral cue; arrow cued trial, where the target appeared at the location indicated by the central arrow cue; and an uncued trial, where the target appeared at remaining two locations. There were two cue validity conditions in the Spatially Convergent case: both cued trial, where the target appeared at the location indicted by both cues; and both uncued trial, where the target appeared at any of the noncued locations. The stimulus onset asynchrony (SOA) separating the presentation of the two cues and the target was 100, 300, 600, or 900 ms. 155 Figure 7.2. Fixation Display 1000ms Spatial ly Divergent Condit ion Spatial ly Convergent Condit ion 7.7% Number Cue (until response) Arrow Cue (until response) SOA 100, 300, 600, or 900ms Target Onset 2600ms number cued arrow cued uncued both cued both uncued Figure 7.2. Illustration of stimuli (not to scale) and sample sequence of events for the predictive central - nonpredictive arrow (PC-NA) cues condition. Every trial began with a 1000 ms presentation of a fixation display followed by a simultaneous presentation of two cues: spatially predictive central number cue (1, 3, 6 or 9) and a spatially nonpredictive central arrow. Central number cues were spatially predictive (with a target appearing at the cued location on 77% of trials), and central arrow cues were spatially nonpredictive (with a target appearing equally often at the cued and uncued location on 7.7% of trials). On any given trial, number direction and arrow direction could diverge spatially (Spatially Divergent) or converge spatially (Spatially Convergent). Both cues were present on the screen for the duration of the trial. There were three cue validity conditions in the Spatially Divergent case: number cued trial, where the target appeared at the location predicted by the central number cue; arrow cued trial, where the target appeared at the location indicated by the arrow cue; and an uncued trial, where the target appeared at remaining two locations. There were two cue validity conditions in the Spatially Convergent case: both cued trial, where the target appeared at the location indicted by both cues together; and both uncued trial, where the target appeared at any of the noncued locations. The stimulus onset asynchrony (SOA) separating the presentation of the two cues Figure 7.3. 156 Fixation Display 1000ms Spatial ly Divergent Condit ion Spatial ly Convergent Condit ion Peripheral Cue (90ms) Number Cue (until response) SOA 100, 300, 600, or 900ms • 7.7% • 9 • 77% 7.7%| 7.7% • • • 9 • • Target Onset 2600ms number cued onset cued uncued both cued both uncued Figure 7.3. Illustration of stimuli (not to scale) and sample sequence for the nonpredictive peripheral - predictive central (NP-PC) cues conditions. Every trial began with a 1000 ms presentation of a fixation display followed by a simultaneous presentation of two cues: a spatially nonpredictive peripheral onset cue and a spatially predictive central number cue (1, 3, 6 or 9). Peripheral cues were spatially nonpredictive (with a target appearing equally often at the cued and uncued location on 7.7% of trials) and central number cues were spatially predictive (with a target appearing at the cued location on 77% of trials). On any given trial, the peripheral onset location and the number direction could diverge spatially (Spatially Divergent) or converge spatially (Spatially Convergent). The peripheral cue was presented for 90 ms while the central number cue was present on the screen for the duration of the trial. There were three cue validity conditions in the Spatially Divergent case: number cued trial, where the target appeared at the location predicted by the central number cue; onset cued trial, where the target appeared at the location indicated by the peripheral cue; and an uncued trial, where the target appeared at remaining two locations. There were two cue validity conditions in the Spatially Convergent case: both cued trial, where the target appeared at the location indicted by both cues together; and both uncued trial, where the target appeared at any of the noncued locations. The Stimulus Onset Asynchrony (SOA) separating the presentation of the two cues and the target was 100, 300, 600, or 900 ms. Figure 7.4 Nonpredictive Periperhal (NP) Cue Effect Nonpredictive Arrow (NA) Cue Effect Predictive Central (PC) Cue Effect 390 370 350 a: ci330 h-310 r -290 NP-NA NP-PC onset cued uncued _L _L _L NP-NA PC-NA arrow cued uncued _L _L _L _L _L _L PC-NA NP-PC — H — number cued uncued _L _L _L _L _L 100 300 600 900 S O A (in ms) 100 300 600 900 S O A (in ms) 100 300 600 900 SOA (in ms) 100 300 600 900 SOA (in ms) 100 300 600 900 S O A (in ms) 100 300 600 900 S O A (in ms) Figure 7.4. Mean RTs for each individual cue type as function of its cue pairing, cue validity and SOA. Left panel shows mean cued and unced RTs as a function of S O A s elicited by nonpredictive peripheral cues (NP) across nonpredictive central arrow (NP-NA) and predictive central (NP-PC) cue conditions. Middle panel shows mean cued and unced RTs as a function of SOA elicited by nonpredictive central arrow cues (NA) across nonpredictive peripheral (NP-NA) and predictive central (PC-NA) cue conditions. Right panel shows mean cued and unced RTs as a function of S O A s elicited by predictive central cues (PC) across nonpredictive central arrow (PC-NA) and nonpredictive peripheral (NP-PC) cue conditions. Ol 158 Figure 7.5. 390 Spatial ly Divergent Spatial ly Convergent 370 h E c 350 h an § 330 h 310 h 290 20 | 16 QE 12 T3 CD zz O ft "O <D ID o c =) arrow cued onset cued uncued — both cued — both uncued 100 300 600 900 S O A ( i n m s ) 100 300 600 900 S O A (in ms) B. Sum of Divergent Effects Convergent Effect Figure 7.5. Mean RTs for the nonpredictive peripheral - nonpredictive arrow (NP-NA) cue condition. Figure 7.5A: Left panel illustrates spatially divergent condition, i.e., where the two cues indicated two different spatial locations and the right panel illustrates results for the spatially convergent condition where the two cues indicated the same spatial location. Mean RTs for arrow cued, onset cued and uncued trials are plotted as a function of S O A for the Spatially divergent case, and mean RTs for both cued and both uncued trials are plotted for the spatially convergent case. Figure 7.5B shows the overall magnitudes of attentional orienting (Uncued RT-Cued RT) for the sum of the two spatially divergent effects and the spatially convergent effect. 159 Figure 7.6. 390 370 E c 350 h-£ 3 3 0 CD CD Spatial ly Divergent Spatial ly Convergent number cued arrow cued uncued both cued both uncued 310 290 I 1 1 L 100 300 600 900 S O A (in ms) 100 300 600 900 S O A (in ms) Sum of Divergent Effects Convergent Effect Figure 7 . 6 . Mean RTs for the predictive central - nonpredictive arrow (PC-NA) cue condition. Figure 7 . 6 A : Left panel illustrates spatially divergent condition, i.e., where the two cues indicated two different spatial locations and the right panel illustrates results for the spatially convergent condition where the two cues indicated the same spatial location. Mean RTs for number cued, arrow cued and uncued trials are plotted as a function of S O A for the Spatially divergent case, and mean RTs for both cued and both uncued trials are plotted for the spatially convergent case. Figure 7 . 5 B shows the overall magnitudes of attentional orienting (Uncued RT-Cued RT) for the sum of the two spatially divergent effects and the spatially convergent effect. 160 Figure 7.7. 100 300 600 900 S O A (in ms) B. 100 300 600 900 S O A (in ms) -oo 16 E H 1 2 h CD 9 8 •o CD O 5 4 Sum of Divergent Effects Convergent Effect Figure 7.7. Mean RTs for the nonpredictive peripheral - predictive central (NP-PC) cue condition. Figure 7.7A: Left panel illustrates spatially divergent condition, i.e., where the two cues indicated two different spatial locations and the right panel illustrates results for the spatially convergent condition where the two cues indicated the same spatial location. Mean RTs for onset cued, number cued and uncued trials are plotted as a function of S O A for the Spatially divergent case, and mean RTs for both cued and both uncued trials are plotted for the spatially convergent case. Figure 7.5B shows the overall magnitudes of attentional orienting (Uncued RT-Cued RT) for the sum of the two spatially divergent effects and the spatially convergent effect. 161 References Berte lson, P. (1967). The time course of preparation. Quarterly Journal of Exper imental Psycho logy, 19, 272-279. Corbetta, M. & Shu lman, G . L. (2002). Control of goal-directed and st imulus-driven attention in the brain. Nature Rev iews Neurosc ience, 3, 201-215. Eimer, M. (1997). Uninformative symbol ic cues may bias visual-spat ial attention: behavioral and electrophysiological ev idence. Biological Psycho logy , 46, 67-71. Findlay, J . M. & Gilchrist, I. D. (2003). Act ive v is ion: The psychology of looking and seeing. Oxford University P ress . Fr iesen, C . K., & Kingstone, A . (2003). Abrupt onsets and gaze direction trigger independent reflexive attentional effects. Cognit ion, 87, B1 -B10 . Fr iesen, C . K., Rist ic, J . , & Kingstone, A . (2004). Attentional effects of counterpredict ive gaze and arrow cues . Journal of Experimental Psycho logy : Human Percept ion & Per formance, 30, 319-329. Jon ides , J . (1981). Voluntary versus automatic control over the mind's eye's movement. In J . B. Long and A . D. Badde ley (Eds.) , Attention & Per formance IX (pp. 187-203). Hi l lsdale, N J : Er lbaum. Kingstone, A . & Pratt, J . (1999). Inhibition of return is composed of attentional and oculomotor p rocesses . Percept ion & Psvchophvs ics , 61 , 1046-1054. Klein, R. M., Kingstone, A . & Pontefract, A . (1992). Orienting of visual attention. In K. Rayner (Ed.), Eye Movements and V isua l Cogni t ion: S c e n e Percept ion and Read ing , (pp. 46-63). North-Hol land: Elsevier Sc ience Publ ishers B. V . Klein, R. M., & Shore , D. I. (2000). Relat ions among modes of v isual orienting. In S . Monsel l and J . Driver (Eds.) , Attention & Performance XVIII (pp. 195-208). Cambr idge, M A : MIT Press . Klein, R. M. (2004). On the control of visual orienting. In M.I. Posner , (Ed.) Cognit ive Neurosc ience of Attention. New York: Guilford P ress . Muller, H. J . , & Rabbitt, P. M. A . (1989). Ref lexive and voluntary orienting of visual attention: time course of activation and resistance to interruption. Journal of Exper imental Psycho logy : Human Percept ion and Per formance, 15, 315-330. Posner , M. (1978). Chronometr ic Explorat ions of the Mind. Hi l lsdale, N J : Er lbaum. Posner , M. I. (1980). Orienting of attention. Quarterly Journal of Exper imental Psycho logy, 32, 3-25. Posner , M. I., & C o h e n , Y . (1984). Components of V isual Orienting. In H. B o u m a & D. G . Bowhuis (Eds.) , Attention & Per formance X (pp. 531-556). Hi l lsdale, N J : Er lbaum. Rist ic, J . , Fr iesen, C . K., & Kingstone, A . (2002). Are eyes spec ia l? It depends on how you look at it. Psychonomic Bulletin & Rev iew, 9, 507-513. Rist ic, J . , & Kingstone, A . (2005). Taking control of reflexive socia l attention. Cognit ion, 94, B 5 5 - B 6 5 . Sternberg, S . (1969). Memory scann ing: Mental p rocesses revealed by reaction time exper iments. Amer ican Scientist, 57, 421-457 Taylor, T. L , & Klein, R. M. (1998). On the causes and effects of inhibition of return. Psychonomic Bulletin & Review, 5, 6 2 5 - 6 4 3 . Tipples, J . (2002). Eye gaze is not unique: Automat ic orienting in response to noninformative arrows. Psychonomic Bulletin & Rev iew, 9, 314-318. V e c e r a , S . P. & R izzo , M. (2006). Eye gaze does not produce reflexive shifts of attention: Ev idence from frontal lobe damage. Neuropsycholoq ia , 44, 150-159. C H A P T E R 8 Genera l D iscuss ion 164 The eleven exper iments that compose the six studies presented in this dissertation have investigated the nature of social attentional orienting, its potential un iqueness relative to other attentional cues, and its relationship with the two classical ly defined modes of attentional orienting — reflexive and volitional attention. Overal l the results indicate that eye direction triggers strongly reflexive attentional effects but that other centrally presented spatially nonpredict ive cues , such as arrow direction, although less strongly reflexive, elicit similar behavioral effects. The results of the study presented in Chapter 3, carried out with individuals with autism and typically developing persons, suggest that socia l reflexive orienting is triggered because humans normally perceive eye direction as conveying key social ly relevant information, such as attentional engagement or disinterest. It is perhaps because orienting to another person's eye direction is so important for everyday socia l communicat ion that social orienting is difficult to inhibit (Chapter 2) and that it is relatively insensit ive to the level of ongoing cognitive activity (Chapter 5) or idiosyncratic cue-target relations (Chapter 6). In other words, the attentional shift that is triggered in response to another individual's gaze is powerful and strongly reflexive. Nevertheless, it is clear that other directional cues , such as arrows, can produce similar behavioral results. The results of Chapter 4 clearly indicated that the standard cuing effects of eye gaze and arrow cues were behavioral ly indist inguishable in both young children and adult observers. Ref lexive orienting of similar t imecourse and magnitude occurred regardless of whether the central cue was a schemat ic face looking left or right or an arrow pointing left or right. However, subsequent examinat ion revealed some key differences between these two effects. In contrast to the strongly reflexive attentional effects elicited by eye direction, the attentional effects elicited by a central arrow cue cannot interrupt an ongoing cognitive activity (Chapter 5) and the effects of 165 an arrow are influenced by arbitrary cue-target cont ingencies (Chapter 6). The distinction drawn between biologically relevant and biologically irrelevant cues exempli f ies the difference between attentional cues that are typically conveyed by people (e.g., eye direction or finger pointing) and those conveyed by inanimate directional objects (e.g., an arrow). The final study in this thesis (Chapter 7) demonstrated that the attentional effects elicited by central spatially nonpredict ive cues operate independently from both the reflexive attention effects elicited by spatially nonpredictive peripheral cues and volitional attention effects elicited by spatially predictive nondirectional central cues . This finding supports a main conclus ion that spatially nonpredictive attentional cues shown at central f ixation, such as eye direction and arrow direction, trigger reflexive shifts of attention that are distinct from both traditionally defined reflexive orienting and volitional orienting. Col lect ively, the f indings of this thesis raise severa l key issues for one 's understanding of socia l attention as well as for understanding of the human attention sys tem, its components, behavioral performance measures , and underlying brain mechan isms. In the following sect ions these issues will be d iscussed in turn. One issue concerns reflexive socia l attention, its properties and putative un iqueness. A second issue concerns the behavioral effect that centrally presented cues have on human attention. The results from the final study in this thesis (Chapter 7) have demonstrated that central spatially nonpredict ive directional cues trigger orienting effects that are independent from the attentional effects elicited by the two c lass ic cue tasks, nonpredict ive peripheral onsets and predictive central cues . Support ing the present conclus ion that central nonpredict ive cues , like eyes and arrows, trigger reflexive orienting, a review of the past literature indicates that these cues meet the standard 166 criteria for 'reflexivity'. This will be demonstrated by a compar ison of central nonpredictive cues and peripheral nonpredictive cues against seven commonly accepted behavioral characterist ics of reflexive attention. A third issue concerns the implications of recognizing central cues as reflexive and what this means for one 's future understanding and conceptual izat ion of reflexive orienting. A fourth issue concerns how the present thesis may have important implications for one 's understanding of volitional orienting. Al though this dissertation did not examine volitional orienting directly, my finding indicating that central spatially nonpredict ive arrows produce reflexive orienting suggests that the results obtained with the c lass ic task using central spatially predictive arrows need to be reevaluated and reinterpreted. Finally, in the future direction sect ion, an attempt will be made to reconci le the present data within the c lass ic framework as well as within a conceptual ly different framework. A n alternative way of understanding and measur ing attention behavioral ly will a lso be presented. Ref lexive Soc ia l Attention F a c e s are significant and biologically highly relevant. Accura te perception and recognition of faces and facial components is critical for successfu l negotiation of the multifaceted socia l environment and interpretation of the complex human social signal ing sys tem. Indeed, numerous animal (e.g., Perrett, 1985) patient (Moscovi tch, Winocur, & Behrmann, 1997), and neuroimaging studies, employing both functional (e.g., Kanwisher, McDermott & Chun , 1997; Kanwisher , 2000) and electrophysiological methods (e.g., Al l ison et al , 1994; Bentin et al , 1996), strongly suggest that perception of faces and their facial features is accompl ished by a distinct and a highly spec ia l ized brain network located,in the temporal lobe of both primates and humans. Th is network for analysis of faces is hypothesized to encompass parts of the fusiform gyrus, involved 167 in detection of faces, and the superior temporal suc lus (STS) , engaged in the analysis of biological motion and eye gaze direction (e.g., Al l ison, P u c e & McCar thy , 2000; Hoffman & Haxby, 2000). Percept ion and accurate interpretation of eye direction of another human appears to be fundamental to healthy social cognit ion, underlying such complex behaviors as turn-taking in conversat ion, as well as enabl ing one to infer the emotional and cognit ive mental states of others (Baron-Cohen, 1995). In fact, as suggested by the results in Chapter 3, an appreciat ion of the social importance of eye gaze s e e m s to be absent in individuals with autism who display profound def ic iencies in socia l functioning. It has a lso been argued that the ability to interpret and follow eye direction is facilitated by the unique morphology of the human eye, which displays a high contrast between the iris and the sc lera (Kobayashi & Kosh ima , 1997). Consistent with this idea, adult humans appear to be remarkably accurate in perceiving fine shifts in gaze position and attributing it to other objects in the environment. For instance, a recent psychophys ica l examinat ion of the perceptual threshold for resolving eye gaze position indicated that accurate discrimination of eye position is achieved when the iris is shifted by approximately 30s/arc (Symons et al , 2004). In other words, people can identify where someone e lse is looking at within 1° of visual angle. G iven this specia l ized role that eye gaze s e e m s to play in socia l cognit ion, the result indicating that eye direction triggers shifts in attention towards the gazed-at location is readily understandable. The present data revealed severa l properties of this social orienting effect. First, the behavioral impact and time course of socia l reflexive orienting is approximately equivalent for children as young as 3 years of age and adults (Chapter 4). This agrees with the results of other developmental studies suggest ing that newborns prefer to scan the eyes of the face (Batki et al , 2002) and that by 12 weeks of 168 age children begin to follow the direction of eye gaze reliably (Hood, Wil len & Driver, 1998; Farroni et al , 2002). Thus , it appears that there is rapid developmental progression regarding socia l reflexive orienting. S e c o n d , reflexive socia l orienting occurs because eye direction conveys social ly relevant information and not because eye direction of another person is typically correlated with important events in the environment (Chapter 3). This distinction is important as it suggests that training individuals with impairments in socia l functioning to utilize eye direction as a predictive social cue may not capture how eye direction is used normally. Finally, the thesis demonstrated in several ways that the strength of the attentional effect that is elicited by eye direction is highly resistant to modulation by cognit ive factors, such as task re levance, ongoing mental activity or the representation of the st imulus (Chapters 6, 5, 2). This agrees with the neuroimaging and patient data indicating that the activation of the S T S is largely unaffected by top-down factors (Vuil leumier et al , 2001), and that the S T S s e e m s to be critical to processing gaze (Hoffman & Haxby, 2000) and is activated when attention is oriented to eye direction (Kingstone, Tipper, Rist ic & Ngan, 2004). Thus the spatial attention shift initiated in response to perceived eye direction appears to be strongly reflexive, as compared to the effects elicited by other directional central cues ; and it is likely to be mediated by unique brain networks responsib le for analys is of eye direction. Ref lexive Attention and Central C u e s Even though social ly relevant eye direction triggers reflexive orienting, the fact remains that other central cues , such as arrows for example, display very similar behavioral effects. Specif ical ly, the investigations presented in Chapters 4-7 indicate that central spatially nonpredict ive arrows produce orienting effects that are often indist inguishable from those produced by eye gaze direction in s imple behavioral tasks. 169 Indeed, the data in Chapter 7 indicated that the reflexive attention effects of arrow cues occurred independently and concurrently with the reflexive effects elicited by peripheral cues and the volitional effects elicited by predictive central cues . The result indicating that the attention effect elicited by arrow cues did not interact with reflexive orienting to peripheral onsets or volitional orienting to a central predictive cue, are consistent with the interpretation that central arrow cues trigger a unique form of reflexive orienting. However it is presently unclear how the behavioral effects of a nonpredict ive central arrow cue compares to five commonly accepted behavioral character ist ics of reflexive orienting, and two criteria of automaticity, which have been establ ished over the past 20 years based largely on the R T patterns of facilitation and inhibition triggered by nonpredictive peripheral cues . The aim of this present sect ion is to compare the reflexive orienting effect triggered by central nonpredict ive arrows against these seven behavioral s tandards. The standard characterist ics of reflexive orienting are as fol lows. The first standard is based on the notion that direct location mapping between the cue and the target, as the peripheral cuing task accompl ishes , is necessary for reflexive orienting to occur (e.g., Posner , Snyder , Dav idson, 1980; Muller & Findlay, 1988). The second is that spatial reflexive orienting occurs in response to stimulation that does not predict where a target is likely to appear. Thus , any benefits in R T that are observed for targets presented at the cued versus noncued locations can be attributed to reflexive orienting because performance was affected by a sensory event to which participants had no incentive to attend to (e.g., Posner , 1980). Third, according to one of the earliest definitions, a physical reaction that occurs less than 100 ms after stimulation constitutes a reflexive response (Posner , 1978). Fourth, when nonpredictive peripheral cues are 170 manipulated in experimental tasks, the early behavioral benefits are typically replaced by a R T cost for targets appear ing at the cued location. B e c a u s e the IOR effect is typically observed after an initial R T facilitation effect in the peripheral cuing task, the IOR effect is commonly considered to represent one of the key behavioral markers of reflexive orienting (Fr iesen & Kingstone, 2003). Fifth, because reflexive attention, as elicited by peripheral cues , has been found to influence select ion of targets presented both within and between sensory modalit ies, (e.g., Ward , 1994), it has been hypothesized that the mechan isms underlying reflexive orienting display supramodal qualit ies. That is, it has been found that attention is oriented reflexively both when the cue and the target match in sensory modality and when they mismatch in sensory modality (e.g., Ward , 1994; Driver & Spence , 1998; McDona ld & Ward , 2000). Addit ionally, the reflexivity of a process is also a s s e s s e d against the two criteria of automaticity. O n e is a load insensitivity criterion, whereby automatic p rocesses display insensitivity to the level of ongoing goal-directed activity, such that the process is not ext inguished or modulated with an increase in the current cognit ive load. The second is the intentionality criterion, whereby automatic p rocesses display insensitivity to voluntary, goal-directed control, such that an automatic process is not ext inguished or modulated by deliberate voluntary control (e.g., Hasher & Zacks , 1978; Yant is & Jon ides , 1990). Below the behavioral effects elicited by nonpredictive peripheral cues and nonpredict ive central arrow cues are compared according to the five characterist ics of reflexive orienting and two criteria of automaticity. C u e Spat ial Posit ion One of the primary dif ferences between peripheral onsets and central arrow cues is their spatial posit ion. Per ipheral stimuli are presented in the parafoveal area of the 171 visual f ield, typically 5 to 8 degrees of visual angle away from central fixation. In contrast, central attentional cues are presented directly at fixation. This difference in cue location is important as , for example, peripheral vision affords relatively poor spatial acuity and it is very sensit ive to st imulus changes in movement or luminance; while foveal vision d isp lays relatively high spatial acuity and permits a fine-grain analys is of the features of the fixated stimuli (e.g., Todd & van Gelder , 1979). A key argument put forward as to why uninformative peripheral stimulation triggers reflexive orienting is that its spatial location provides information directly about a possible target location. A s such , these direct cues do not require any significant cognit ive interpretation, i.e., they simply "tag" directly a location in visual space . In contrast, a key argument put forward as to why spatially informative central cues engage volitional orienting is that these attentional cues , such as left- and right-pointing arrows, require a relatively high-level of cognitive interpretation before their spatial meaning can be determined. This dichotomy, which is reflected by the two c lass ic cuing tasks, was perhaps solidified by the seminal paper publ ished by Jon ides in 1981. In his study subjects were asked to search a briefly presented array of letters for the target letter (L or R) . Before the array appeared, an arrow cue, either randomly pointing to one of the target locations or reliably indicating one of the target locations, was f lashed momentari ly either at fixation or at a peripheral location. If subjects were told to ignore the central arrow, orienting to the cued location was absent compared to the condition where subjects were told to utilize the central arrow. This difference suggested to Jon ides (and many others) that a nonpredict ive central arrow cue does not trigger reflexive orienting. In contrast, a peripheral arrow cue a lways triggered a shift in attention to the cued location regardless of whether the participants were told to ignore the peripheral arrow cue or not. B a s e d on these data Jon ides (1981) conc luded that a 172 peripheral cue is more effective in attracting attentional resources than a central cue, because the latter requires encoding and an analys is of the cue 's meaning before a shift of attention is initiated. However, a c loser inspection of the literature suggests that this division based on spatial position may be quite superf icial . Indeed, the studies investigating the utility of attentional cues presented at peripheral and central locations indicate that both cues cause a shift in attention only when they are interpreted as being meaningful . Thus , it is the meaning attached to this transient change in the display, which is also reflected by the peripheral cuing task, that is responsible for reflexive orienting toward a peripherally cued location. The results indicating that a peripheral onset st imulus draws reflexive attention to its location only when it s ignals an appearance of a new object support this conclus ion (e.g., Yant is , 1993; Yant is & Hillstrom, 1994; Yant is & Jon ides , 1996; Yant is & Egeth, 1999; Enns et a l , 2001). That is if the visual transient in the display (i.e., st imulus motion or luminance change) does not signal the appearance of a new object it a lso fails to capture attention reflexively. W h e n these findings are extended to the attentional cuing literature, which postulates that luminance increments summon reflexive attention to the cued location, it appears that this holds true only when participants adopt a particular attentional set which is sensit ive to detection of new objects (e.g., Folk, Remington & Johnston 1992, Yant is & Hil lstrom, 1994). Hence , collectively these data strongly suggest that a peripheral luminance transient triggers a shift of attention to the cued location because it s ignals an appearance of a new object in visual space . The attentional cuing effect therefore s e e m s to be largely dependent to the top-down attentional set adopted by the observers. Thus it appears that while peripheral attentional cue may stimulate the spatial location directly, the attentional orienting effect is dependant on the meaning that participants attach to that cue. 173 Similarly, central arrow cues are thought to shift attention reflexively only after adequate time is al lowed for cue process ing. Al though the result reported by Jon ides (1981), indicating that central cues do not trigger reflexive orienting of attention was taken as supporting ev idence that peripheral cues are more efficient than central cues in triggering attentional orienting, this finding has not been replicated by severa l more recent investigations. In contrast to Jon ides ' data, current studies (e.g., Eimer, 1997; Hommel et al , 2001 ; Rist ic, Fr iesen & Kingstone, 2002 (Chapter 4); T ipples, 2002) indicate that spatially uninformative central arrows trigger a shift in spatial attention to the pointed-at location. G ibson and Bryant (2005) recently addressed this d iscrepancy between the c lass ic study by Jon ides (1981) and the more recent f indings, and conc luded that the recent nonreplications of Jon ides ' data may be due to key, methodological di f ferences. Specif ical ly, in the Jon ides ' study, arrow cues and target d isplays were presented very briefly, for 25 ms, and with a very brief cue-target stimulus onset asynchrony ( S O A ) of 50 ms while in the more recent investigations arrow cues are presented for a longer period of time, anywhere between 75ms up to the length of the trial and target d isplays routinely remain on once they are presented. G ibson and Bryant (2005) adapted Jon ides ' methodology and included both short (25ms) and longer (200ms) arrow cue durations. The authors observed that the attentional orienting effect did not emerge when both the cue and the cue-target S O A s were very short (25 and 50 ms respectively), replicating the original Jon ides ' (1981) study. However, this null f inding did not hold true for other cue durations and S O A condit ions, in which orienting effects triggered by spatially nonpredictive central arrows were observed for both short and long arrow cue durations. Thus the emergence of the reflexive orienting triggered by central arrow cues may depend critically on the extent to which the cue has been processed (Gibson & Bryant, 2005), suggest ing that orienting triggered by central 174 nonpredictive cues proceeds only when sufficient time is al lowed for the cues ' attributes to be p rocessed. Therefore, it appears then that while peripheral and central cues may occupy different spatial posit ions, the attentional effects elicited by both cues are not devoid of cognitive inf luences. In the case of peripheral onset cues , studies on attentional capture suggest that visual transients capture attention reflexively only when they signal the appearance of a new visual object. Similarly, in the case of central arrow cues , sufficient cue exposure time is necessary for the attentional orienting to occur, thus permitting an interpretation of cue 's shape and its meaning (Gibson & Bryant, 2005). Thus , both reflexive attention effects triggered by peripheral and central cues appear to rely fundamental ly on meaning-based cue interpretations. C u e Information Ref lexive behavioral p rocesses are assumed to ar ise in situations when the response occurs independently from an observer 's immediate goals and expectat ions. Captur ing this important idea, attentional cues chosen to engage reflexive attention are spatially uninformative with regard to a forthcoming target's spatial posit ion, i.e., because the cue does not predict where a target is likely to appear, an observer should not generate an expectancy that the target will appear at the cued location. Hence, any spatial orienting effects of a cue can be attributed to reflexive p rocesses . Experimental ly, in a cuing task where a target can appear at one of two possible locations, this means that half the time the target occurs at the cued location and half the time the target occurs at the uncued location. Part icipants are typically informed about this random pairing between the cue and the target, and more often than not the participants are instructed to ignore the attentional cue (e.g., Jon ides , 1981; Posner , 1980). Despite this explicit forewarning with regard to the arbitrary cont ingency 175 between the cue and the target, a cue that attracts attention reflexively d isplays behavioral facilitation at the cued location relative to a noncued location. Consis tent with this logic, both spatially nonpredict ive peripheral cues and spatially nonpredict ive central cues can be said to trigger reflexive shifts of attention toward the cued location, insofar as they result in a processing facilitation at the cued location when participants are informed that the cue does not, in any way, predict the subsequent location of the target. In a seminal study, Posner (1980) asked participants to detect targets occurring randomly either at the location of a spatially nonpredict ive peripheral cue or at the mirror opposite location. The result was that target detection time was facilitated at the cued target location compared to the uncued target location despite the fact that the luminance onset did not indicate reliably where the target would appear. Numerous other studies conducted s ince then have replicated this behavioral result (e.g., Jon ides, 1981; Posne r & C o h e n , 1984; Br iand & Klein, 1987; Muller & Rabbitt, 1989; Rafal et al , 1989; Pratt & Ab rams , 1995; Reuter-Lorenz, J h a & Rosenquis t , 1996; Danzinger & Kingstone, 1999). Similarly, in an early demonstrat ion of reflexive orienting elicited by a central cues , Fr iesen and Kingstone (1998) asked participants to local ize, identify and detect targets randomly occurring either at the location gazed-at by a central schemat ic face or at the opposite location. A s noted previously, the results indicated that R T was facilitated for targets appear ing at the cued (gazed-at) location compared to the uncued (not gazed-at) location. Subsequent investigations, including those in the present thesis, extended this result indicating comparable results occurr ing for other directional cues , like head direction, finger pointing, and arrow direction (e.g., Hood, Wil len and Driver, 1998; Driver et al 1999; Langton & Bruce, 1999; Hommel et a l , 2001; Rist ic et al , 2002; 176 Tipples et al , 2002). In sum, like the tasks using peripheral onset cues , there is now a wealth of studies reporting that central spatially nonpredictive cues trigger reflexive orienting when cue direction and target location are randomly assoc ia ted . Early Facil itation A fundamental characterist ic of an attentional reflex, in addition to its emergence in the absence of any del iberate intent, is that it should occur rapidly, within 100 ms or less after the initial appearance of the stimulating event (Posner, 1978). The cuing effect for peripheral cues satisf ies this characterist ic of reflexive orienting. Importantly, however, so does the cuing effect for nonpredictive central cues (e.g., Fr iesen & Kingstone, 1998; Rist ic et al, 2002; as well as thesis chapters 4, 5, 7). The rapid effects of early R T facilitation observed for peripheral and central cues is corroborated by electrophysiological ev idence, in which the early attention-directing E R P waveform is manifested as increased amplitude of the first positive deflection, the P1 component , peaking at about 90-140ms after cue presentation (e.g., Luck, W o o d m a n & Voge l , 2000). Overal l , the E R P data indicate that both spatially nonpredictive peripheral onset cues (Luck et al , 1994) as well as nonpredict ive central gaze and arrow cues (Eimer, 1997; Schul ler & Ross ion , 2001 ; Schul ler & Ross ion , 2004) trigger P1 attention effects as early as 100 ms after the stimulus onset. Inhibition of Return (IOR) One of the most critical and most often cited dif ferences between the effects of peripheral and central attentional cues is that the early facilitation effect for peripheral cues is replaced by a later inhibitory effect; whereas the early facilitation effect for central directional cues persists for up to a second or more, with the IOR effect rarely, if ever, emerging (see Fr ischen & Tipper, 2004). In the first experimental demonstrat ion of the IOR effect, Posner and C o h e n (1984) observed that, in the peripheral cuing task, as 177 the cue-target onset delay interval increased beyond 300 ms, participants' R T was longer for targets appear ing at the cued locations compared to uncued target locations. Denoting the idea that reflexive attention was inhibited in returning to previously attended locations, Posner and C o h e n (1984) named this effect the "inhibition of return (IOR) effect. Because the IOR effect was typically observed following early facilitation to a peripherally cued location, it has come to be considered as one of the key behavioral markers of reflexive orienting (e.g., Fr iesen & Kingstone, 2003). In contrast, investigations that have used central nonpredictive cues typically fail to observe the IOR effect (e.g., Fr iesen & Kingstone, 1998; Langton & Bruce 1999; Driver et al , 1999; Rist ic et a l , 2002; Tipples, 2002; Langton, 2000; Hommel et a l , 2001; Fr iesen et a l , 2004; but see Fr ischen & Tipper, 2004), with the R T effects displaying shorter R T s for targets appear ing at cued compared to uncued target locations at short and long cue-target intervals. One of the first theoretical accounts of the IOR effect postulated that it occurs as a consequence of reflexive attention being oriented away from the cued location, the result being that search efficiency of the environment is enhanced by inhibiting reorienting to recently attended locations (e.g., Klein, 1988). Therefore, IOR was considered to arise as a consequence of reflexive orienting of attention. This notion predicts that IOR should a lways be preceded by a reflexive shift of attention. This conclus ion is not, however, supported by the data. Severa l l ines of ev idence indicate that an IOR effect can be triggered independent of, or concurrent with, a shift in reflexive attention. For example, when the attended location and to-be inhibited location are spatially d issociated (these two locations are typically confounded in the c lass ic cuing paradigm with two possible target locations) facilitation at the attended location and IOR at the to-be-inhibited location can co-occur at an S O A as brief as 50 ms 178 (Danziger & Kingstone, 1999; see a lso Maruff et al , 1999). Dovetail ing with this finding that reflexive orienting and IOR are d issociable, Fr iesen and Kingstone (2003) demonstrated independent orienting effects of reflexive attention and IOR using eye direction cues . In their experiment, a schemat ic face cue served both as a spatial nonpredict ive eye g a z e cue and as an abrupt onset cue. They reported that the detection of targets presented at the gazed-at locations was facilitated at the same time that detection of targets presented at the abrupt onset location was inhibited, with the magnitude of the IOR effect not varying as a function of the observed facilitation effect. If IOR does not occur as a consequence of a reflexive attention being drawn to the cued location, why does it occur? A considerable amount of ev idence now indicates that IOR occurs as a consequence of oculomotor inhibition (see Taylor & Kle in, 1998; and Kle in, 2004 for reviews; although it should be noted that whether the entire IOR effect can be attributed to oculomotor p rocesses is still an issue of considerable d iscuss ion with severa l recent studies suggest ing an attentional role the IOR effect (e.g., Pr ime & Ward , 2004; Snyder & Kingstone, in press). A n abrupt onset in the visual periphery can reflexively capture both attention and the eyes (Theeuwes, et al , 1998). However, this does not demand the conclus ion that both the attention and the eyes are attracted to salient external events because the two forms of orienting are linked by a common neural architecture as it could a lso be the case that each is independently activated by an abrupt onset. Indeed, it appears that the latter is the case , with attentional attraction leading to a R T facilitation effect, and eye movement activation leading to IOR (Klein, 2004; Hunt & Kingstone 2003a; 2003b). In one of the c lass ic demonstrat ions of this notion Rafal et al (1989) varied both the type of the cue presented (peripheral vs . central) with an instruction to execute or withhold eye movements (eye movements vs . no eye movements) . Their results indicated that the 179 IOR effect was observed in all condit ions in which an eye movement was generated or programmed regardless of whether the cue was peripheral or central cued , suggest ing that inhibition does not ar ise as a consequence of attentional orienting towards the cued location but largely (if not exclusively) as a consequence of occulomotor activation. In summary, when one examines the important issue of the relationship between the IOR effect and reflexive orienting triggered by peripheral and central cues , it is apparent that both attentional cues display a similar relation to IOR. For instance, both central and peripheral cues produce concurrent effects of longer lasting facilitation and IOR when to-be attended and to-be-inhibited locations are separated spatial ly and when the involvement of the oculomotor sys tem is minimized (Danziger & Kingstone, 1999; Fr iesen & Kingstone, 2003). Supramoda l Ref lexive Orienting. Recent c rossmoda l research strongly suggests that, in addition to influencing the select ion of stimuli within the s a m e sensory modality, attentional sys tem may have a substantial influence on the select ion of stimuli presented in different sensory modalit ies. One of the most fundamental quest ions relating to attentional orienting across sensory modalit ies concerns the issue of whether reflexive attention is modality specif ic or whether it d isp lays supramodal qualit ies, such that its effects are evident across sensory modali t ies. Adopt ing the experimental methods for studying attention in the visual domain , c rossmodal links in reflexive attention orienting have typically been investigated using the attention cuing paradigm (e.g., S p e n c e & Driver, 1994; Ward , 1994; S p e n c e & Driver, 1997; McDona ld & Ward , 2000). In this variant of the c lass ic cuing task, a subject 's attention is typically attracted towards a position in space using an attentional cue (e.g., a visual peripheral onset cue). The target, demanding either a 180 detection or localization response, is then presented either in the same sensory modality, (e.g., visual) or in different sensory modality (e.g., auditory). The results indicate that, with very few except ions (e.g., Ward , 1994; Spence & Driver, 1997), a nonpredictive attentional cue presented in one sensory modality facilitates processing of targets presented in a different sensory modality. For example, in one of the earliest experimental demonstrat ions of reflexive c rossmodal attentional modulat ion, Ward (1994) investigated the effects of nonpredictive auditory and visual attentional cues on localization of auditory and visual targets. Auditory cues triggered attentional orienting to targets presented in the auditory modality whereas visual cues triggered attentional orienting to targets presented in both the visual and auditory domain. Extending these findings to include other sensory modalit ies, S p e n c e et al (1998) investigated the attentional effects of vision, audit ion, and touch in a ser ies of three exper iments. In this study participants were asked to either discriminate tactile targets after being cued by a spatially nonpredictive tone or a f lash, or to discriminate auditory and visual targets after being cued by a spatially nonpredictive tactile cue. Their results indicated that spatially uninformative audio-visual cues triggered orienting toward tactile targets. Similarly, spatially uninformative tactile cues triggered reflexive orienting toward targets presented in both visual and auditory modalit ies. Thus , the results indicate that reflexive orienting proceeds crossmodal ly such that attentional cues presented in one modality (e.g., v isual , auditory, or tactile) trigger orienting towards events presented in another modality (e.g., v isual , auditory, or tactile). And of critical importance here, reflexive attentional shifts triggered by spatially nonpredictive peripheral v isual cues facilitate behavioral responses towards targets presented in other sensory modalit ies (e.g., Ward , 1994; S p e n c e et a l , 1998; Ward , McDona ld & Lin, 2000; McDona ld et a l , 2001). 181 Similarly, a recent study by So to -Faraco et al (2005) demonstrated that a central nonpredict ive eye gaze cue inf luenced allocation of reflexive attention to tactile targets. Part icipants saw a schemat ic face at central fixation, gaz ing either to the left or to the right. Part icipants were asked to detect or discriminate a vibrotactile stimulation del ivered to the left or right finger, occurr ing 100, 300 or 1000 ms after the presentation of the central face cue. Data from speeded detection, speeded discrimination and signal detection tasks converged on the conclusion that a nonpredictive central gaze cue triggered reflexive orienting which facilitated the processing of tactile stimuli at the gazed-at location. This result was replicated with a nonpredict ive arrow st imulus. Thus the Soto-Faraco et al (2005) study provides strong ev idence that the attentional effects of central attentional cues , similarly to the attentional effects of peripheral attentional cues , trigger supramodal shifts of covert reflexive spatial attention. Criteria for Automaticity In the 1980s, considerable research effort was devoted to understanding and developing criteria for the definition of automatic cognit ive p rocesses (e.g., Jon ides , 1981; Hasher & Z a c k s , 1984; Jon ides e t a l , 1985, Naveh-Ben jamin & Jon ides , 1986; Pa lmer & Jon ides , 1988). Two general features of automatic processing emerged. First, because automatic processing should require little cognit ive resources, its effects should be unaffected by manipulat ions that change (e.g., increase) the demands for cognitive resources. S e c o n d , because automatic p rocesses are thought to occur as a result of automatic pathway activation, their emergence and the magnitude of their effect should not be influenced by the amount of top-down voluntary cognit ive control exerted by individuals (Palmer & Jon ides , 1988). These two criteria have been coined the load insensitivity criterion and the intentionality criterion of automaticity, respectively. 182 The two criteria have been operat ional ized in experimental manipulat ions in a number of different ways . For example, load insensitivity has been tested by manipulating the amount of cognitive resources required by a secondary task and the effects these manipulat ions have on performance of a primary, putatively automatic, task (e.g., Jon ides , 1981). Alternatively, load insensitivity has been tested by assess ing whether reflexive orienting interrupts a goal directed activity (e.g., Muller & Rabbitt, 1989), or by assess ing whether the emergence of reflexive orienting is affected by changes in st imulus presentation f requency (e.g., Jon ides , 1981). L ikewise, the influence of voluntary control over automatic p rocesses , or the intentionality criterion, has been tested in manipulat ions examining, for example, the c i rcumstances under which reflexive process may be resistant to suppress ion (e.g., Jon ides , 1981; Yant is & Jon ides , 1990). Evaluat ion of the literature on reflexive orienting of spatial attention, elicited both by traditional peripheral onsets and nonpredictive central cues indicate the attentional effects triggered by these two cues both satisfy the load insensitivity criterion and, similarly, both fail to satisfy the intentionality criterion. These two issues are d iscussed below. Load Insensitivity Criterion. In attentional cuing tasks the load insensitivity criterion has typically been investigated using a secondary task procedure, manipulating the demands for cognitive resources (Jonides, 1981) or using a counterpredict ive cuing task procedure that sets reflexive and volitional orienting in direct competit ion (Muller&Rabbit , 1989; Fr iesen, Rist ic & Kingstone, 2004, Chapter 5). Returning to the seminal study by Jon ides (1981), in his second experiment participants were asked to perform both a target search task and a secondary memory span task. Attentional orienting elicited by peripheral cues was relatively unimpaired by this secondary 183 manipulation whereas attentional orienting elicited by central cues was significantly compromised. These data suggested that reflexive orienting is far less dependant on cognitive resources than volitional orienting. However, there is ev idence that central and peripheral cues behave similarly in cuing tasks that test the load insensitivity criterion by pitting the effects of reflexive and volitional attention directly against one another. In 1989, Muller and Rabbitt publ ished the now c lass ic study, which tested the load insensitivity criterion for reflexive orienting elicited by peripheral cues . In their task observers were cued by a central arrow, which correctly indicated a target's position on 7 5 % of trials. Before the target was presented, a task irrelevant peripheral luminance cue appeared randomly in one of the possible target locations. The critical quest ion was whether the appearance of the task irrelevant stimulus cue at an unlikely target location would disrupt a voluntary act of attentional orienting towards the likely target location. Muller and Rabbitt (1989) found that presenting an irrelevant st imulus at an unlikely location interrupted volitional orienting to the predicted target location, demonstrat ing that reflexive orienting triggered by peripheral cues satisf ies the load insensitivity criterion of automaticity. In a comparable study that examined the load insensitivity criterion for orienting to central nonpredictive cues , Fr iesen, Rist ic and Kingstone (2004, Chapter 5; Experiment 1; see also Driver et al , 1999) employed a so-cal led counterpredict ive cuing paradigm. Part icipants were presented with a central schemat ic gaze cue, the direction of which indicated that the target was likely to appear at the opposite spatial location. In the majority of all the trials (i.e., 80%) the cue correctly indicated the target position, but in the remaining trials (20%), the target appeared at some other possible location. The key compar ison of interest was again whether targets that appeared at the cued location would be detected most quickly despite the fact that participants were 184 instructed to voluntarily al locate their attention towards the opposite, likely target location. Fr iesen et al (2004) observed that central eye direction cues triggered reflexive orienting to the cued location indicating that similarly to peripheral onset cues , reflexive orienting triggered by central eye direction satisf ied the load insensitivity criterion of automaticity in that reflexive orienting towards the unlikely target location w a s observed despite the explicit instruction and act of orienting attention volitionally to a different location. Interestingly, Fr iesen et al (2004) observed that not all central cues are al ike in their ability to trigger attention reflexively. In their Experiment 2, Fr iesen et al (2004) used counterpredict ive central arrow cues instead of central eye gaze cues . These data demonstrated a nonsignif icant trend towards reflexive orienting of attention toward the cued target location at an early S O A of 100 ms and as such did not strictly replicate the early reflexive effects obtained with eye direction stimuli (Experiment 1). Therefore, while in general central attentional cues display behaviorally indist inguishable effects in simple R T tasks, there appear to be differences between the individual central cues in terms of their strength to attract attention reflexively (see also Chapter 6). In sum, the research ev idence suggests that reflexive orienting triggered by both spatially nonpredict ive peripheral and central cues can interrupt or co-occur with volitional orienting in experimental tasks that contrast the effects of reflexive and volitional attention directly. Therefore, reflexive orienting triggered by both peripheral and central attentional cues appears to satisfy the load insensitivity criterion. Intentionality Criterion. The second criterion of automaticity, the intentionality criterion, has been operat ional ized by tasks that require subjects to deliberately suppress orienting to irrelevant stimuli. The first support for the intentionality criterion comes from the results indicating that reflexive attentional orienting is observed despite 185 the instructions that the cues are spatially nonpredict ive. More stringent tests of the intentionality criterion include experimental manipulat ions examining the effects of increased task practice (Warner, Juo la & Kosh ino, 1990) and the effects of deliberate voluntary control (e.g., Yant is & Jon ides , 1990). Thus the critical quest ion of interest is whether the effects of a reflexive, either peripheral or central cue, can be overr idden? Resea rch ev idence suggests that, for both types of the cue, the answer is yes . That is, it appears that both cues violate intentionality criterion similarly in that the reflexive effects can be abol ished by top-down volitional control. Severa l l ines of ev idence indicate that reflexive orienting in response to peripheral cues does not satisfy the intentionality criterion of automaticity. First, it appears that a peripheral cue 's ability to interrupt voluntary orienting diminishes with increased practice. Warner et al (1990) trained participants to orient away from the location indicated by a peripheral cue and demonstrated that, in contrast to Muller and Rabbitt (1989), after extensive training, an infrequently occurr ing peripheral cue c e a s e d to interrupt voluntary orienting of attention. S e c o n d , at least two other studies indicated that an irrelevant peripheral transient failed to capture attention when subjects had no reason to attend to the onset event at all. Yant is and Jon ides (1990, Exper iment 2) asked subjects to discriminate the target (E or H) among distractor letters arranged in along vert ices of a hexagon. Before a sudden onset of the distractor, an arrow cue, indicating the location of a subsequent target with 100% certainty appeared at the center of the sc reen. This perfectly reliable information conveyed by the arrow cue effectively el iminated any attentional effects of the peripheral onset. The s a m e result was replicated in Experiment 3, which manipulated the reliability of the central arrow cue. The data obtained with 100%, 7 5 % and 2 0 % valid central cues indicated that while 186 a completely reliable cue, i.e., 100% valid, el iminated any reflexive orienting towards the onset location, partially reliable cues did not, suggest ing that the effects of peripheral onset cues can be overr idden when the observer has no intention of attending to the onset cue (cf. Theeuwes , 1991). Like peripheral cues , reflexive orienting in response to central nonpredict ive cues , strictly speak ing, a lso fails to satisfy the intentionality criterion of automaticity. Rist ic and Kingstone (2005, Chapter 2) examined whether reflexive orienting triggered by spatially nonpredictive eye direction proceeds in a purely automatic fashion, such that it is completely insensit ive to top-down modulat ion. Part icipants in all condit ions were shown an identical ambiguous central st imulus, containing a left or right pointing eye gaze cue that could be interpreted either as representing "eyes" or a "car". The critical variable of interest was whether this top-down representation of the central cue would in any way modulate reflexive orienting triggered by a spatially nonpredictive eye direction. The results indicated that the emergence of reflexive attention triggered by eye direction is inf luenced by top-down mechan isms, in that the physical ly identical directional cue failed to trigger orienting effects when the observers, adopted an irrelevant cognitive interpretation of the cue 's meaning. However, after the participants adopted the relevant interpretation of the cue (i.e., eyes) its effects could no longer be overr idden. Thus , the ev idence indicates that reflexive orienting triggered by both spatially nonpredict ive peripheral and central cues fails to satisfy the intentionality criterion of automaticity. That is, the attention effects of both peripheral and central cues can be inf luenced by top-down factors. Peripheral onset cues do not a lways capture attention in the absence of current goals and intentions of an individual, and similarly, the emergence of the orienting effect triggered by central attentional cues s e e m s to be 187 critically inf luenced by the top-down representation adopted by the observers . Interestingly, in a direct contrast between peripheral and central cues , it would appear that eyes are if anything "more automatic" than peripheral onsets, in that their effects could not be overr idden once observers represented the ambiguous st imulus in Chapter 2 (Ristic & Kingstone, 2005) as depicting eyes . Summary The outcome of this analysis has shown that the attentional effects triggered by spatially nonpredictive peripheral and central attentional cues produce behavioral ly effects that are extremely similar. Both cues trigger orienting effects (1) that depend on the meaning of the cue despite the difference in their spatial position at the time of presentation; (2) even when the relationship between the cue 's posit ion and the target position is random; (3) that emerge rapidly, by 100 ms after the cue presentat ion; (4) that are long lasting and not accompan ied by inhibition of return when the involvement of eye movement sys tem is minimized; and (5) that are evident both within and across sensory modalit ies. W h e n the orienting effects produced by the two types of cues are a s s e s s e d against the two criteria for automaticity, the results indicate that both nonpredict ive peripheral and central attentional cues trigger reflexive effects that have an ability to interrupt or co-occur with an ongoing cognitive activity but that the attentional effects of both peripheral and central attentional nevertheless can be modulated intentionally by top-down processes . Consequent ly , it follows that insofar as orienting to peripheral abrupt onsets is conceptual ized as being reflexive, as conventional w isdom maintains (e.g., Jon ides , 1981; Muller& Rabbit , 1989), for the same reasons, orienting in response to central cues must a lso be conceptual ized as being reflexive. 188 Implications for Understanding Ref lexive Orienting Accord ing to Kahneman and Tre isman (1984), who advanced a more f ine-grained distinction between automatic p rocesses , automatic p rocesses are hypothesized to vary in strength according to whether one or both of load insensitivity or intentionality criteria are satisf ied. Strongly automatic p rocesses a lways satisfy both criteria, weakly automatic p rocesses satisfy one of the two criteria, and non-automatic p rocesses never satisfy either of the two criteria. A range of spatially nonpredict ive central attentional cues , including eye direction, arrow direction, head deviat ion, finger pointing, words with spatial meaning and even numbers, produces R T facilitation for targets appear ing at a cued location. This broad result is stable and well replicated across the different types of central cues . A s previously suggested, this behavioral effect is best conceptual ized as an instance of reflexive attentional orienting as central nonpredictive cues meet the five characterist ics of reflexive orienting and two criteria of automaticity that have been appl ied to the traditional form of reflexive orienting observed for peripheral onset stimuli. Within this general domain of central stimuli that can trigger reflexive orienting, it does appear that some cues produce effects that are more strongly reflexive than others. At the strongly reflexive extreme, as reviewed above, there appear to be abrupt onsets and gaze cues . Arrows however, s e e m to be less reflexive than eyes . A s indicated by Fr iesen, Rist ic and Kingstone (2004, Chapter 5), eyes but not arrows trigger an attention shift to the cued location even when the goal of participants is to direct attention volitionally to a different location. And as shown in Chapter 6, the reflexive orienting effect for gaze is unaffected by arbitrary cue-target cont ingency relationships. In contrast, arrow cues are affected by such cont ingencies (see also Pratt and Hommel , 2003). 189 At the other end of this cont inuum of reflexivity there resides the reflexive orienting effect produced by spatially nonpredictive central number cues . F ischer et a l . (2003) reported that spatially nonpredict ive central digit cues produce reflexive shifts of attention as if representing numbers on a mental number line running from left to right. In their study, participants were fastest to detect a target appearing on the left side of the screen when the central number cue was numerical ly low (1 or 2), and participants were fastest to detect targets appear ing on the right side of the screen when the central number cue was numerical ly high (8 or 9). However, unlike the effect for abrupt onsets, eyes and arrows, this effect was relatively s low to emerge suggest ing that it may not be as strongly reflexive at cues like onsets , eyes and arrows. Two recent studies have confirmed this hypothesis. Rist ic, Wright and Kingstone (in press) have reported that merely adopting a simple mental set abol ishes the proposed reflexive effect of the number mental line. After replicating the basic finding of F ischer et a l . (2001) — with faster R T s for targets on the left when preceded by a low number cue and faster R T s for targets on the right when preceded by a high number cue — Rist ic et al . asked participants to imagine a number line running from right to left, e.g. , low numbers on the right and high numbers on the left. The result was a profound reversal of the F ischer et al . f inding. Now, R T s were faster for targets on the right when preceded by a low number cue and faster for targets on the left when preceded by a high number cue. In a follow-up experiment, participants were asked to imagine a clock face and targets could appear above, below, to the left or to the right of center. The results indicated that the highest number cue (12) resulted in targets being detected most quickly at the 12 o'clock posit ion, and similarly, the number cues 3, 6, and 9 resulted in targets being detected most quickly at the 3, 6, and 9 o'clock position. This result emerged despite the fact that the central 190 number cues did not predict where a target was going to appear. In a similar study Gal fano et a l . (in press) have shown that when participants are asked to orient attention volitionally in response to a low or high number cue, with the number 1 predicting a target on the right and the number 9 predicting a target on the left, there is no ev idence of reflexive orienting as suggested by F ischer et al . (2003). Col lect ively these two studies indicate that reflexive attentional orienting elicited by number cues occurs under very limited c i rcumstances and is extremely vulnerable to changes in mental set. For instance, the number line effect is determined by the current cognitive top-down representation (Ristic et a l . , in press) and it is unable to interrupt volitional orienting (Gal fano et a l , in press). Taken together these data argue against the notion of reflexive orienting as an al l-or-none phenomenon. A s demonstrated by the data presented in this dissertat ion, as well as by other recent behavioral investigations, reflexive orienting appears to exist on a cont inuum from strongly reflexive (such as eye direction) to weakly reflexive (such as number cues) . Implications for Understanding Voli t ional Orienting In contrast to reflexive attention, voluntary orienting is conceptual ized as arising from the consc ious al location of attentional resources by an observer towards sensory events of interest. In the central arrow cuing task that is typically used to measure voluntary orienting, the fundamental volitional characterist ics of spatial attentional orienting are thought to be revealed by requiring subjects to detect a target light at a peripheral location that is, or is not, pointed at by a central arrow. Importantly, in this task a central arrow cue predicts where a target stimulus is likely to appear. B e c a u s e the spatial effects of a central arrow are assumed to occur only when the arrow is 191 informative as to where a target is likely to appear, its attention effects are attributed to volitional orienting of spatial attention (e.g., Jon ides , 1981). Of course, the studies in this thesis have shown that the above assumpt ion is incorrect. That is, shifts of attention to a cued location will occur even when a central arrow cue does not predict where a target is likely to appear. This finding raises the following issue: If spatially nonpredict ive central cues , such as arrows, trigger a reflexive shift in spatial attention to the cued location, then what has the traditional spatially predictive central arrow cuing task been measur ing? There are several possibi l i t ies. One is that the many past studies that used a predictive arrow cue to study volitional attention, have been measur ing reflexive attention rather than volitional attention. A second possibil ity is that previous investigations have been measur ing volitional attention correctly when a central arrow cue is spatially predictive, i.e., only volitional orienting is engaged when the arrow cue is predictive. A third possibil ity is that a central spatially predictive arrow engages both reflexive and volitional attention, with these effects combining in an additive fashion. A fourth possibil ity is that a central spatially predictive arrow engages both reflexive and volitional attention, with these effects combining in an interactive manner. To distinguish between these four alternatives, in a recent study, Rist ic and Kingstone (in press) used a central cue that does not trigger reflexive shifts of attention when it is spatially nonpredictive, and does engage volitional attention when it is spatially predictive. These two effects were compared against the reflexive attentional effect of an arrow cue when it is spatially nonpredictive and the attentional effect of an arrow cue when it is spatially predictive. In this way, Rist ic and Kingstone were able to determine whether a central predictive arrow engages : (i) only reflexive attention; (ii) only volitional attention; (iii) the summat ion of reflexive and volitional attention; or (iv) the 192 interaction of reflexive and volitional attention. The results indicated that the magnitude of the orienting effect produced by a spatially predictive arrow cue a lways exceeded the magnitude of the individual reflexive attention and volitional attention effects, as well as the sum of the reflexive and volitional components. That is, the data indicated that the orienting effect that is generated by a predictive arrow reflects an interaction between reflexive and volitional attention, and not volitional attention in isolation as has been assumed in the past (e.g., Jon ides , 1981). This point cuts across all levels of behavioral investigation that have used the predictive central arrow cuing paradigm to measure volitional orienting. From those studies focused on discover ing the t ime-course of voluntary orienting (e.g., Jon ides , 1980; Jon ides , 1981; Mul ler & Rabbit , 1989; Muller & Humphreys, 1991), to those that sought to understand the effects of volitional attention on response time and response accuracy (e.g., Posner , et al 1980; Chea l & Lyon, 1991), to those that attempted to understand the effect of volitional attention on perceptual sensitivity and response bias (e.g., Muller & Findlay, 1988; Hawkings et a l , 1990), to those that compared space -based and object-based attention (e.g., Egly, Driver & Rafa l , 1994), to recent attempts to develop a single test to evaluate human attentional networks (e.g., Fan et al 2002). It would appear that these and many studies like them concluded incorrectly that they had engaged and measured volitional orienting. A similar limitation is found when one considers studies that have sought to examine def ic iencies in volitional attention and its neural underpinnings. Difficulties in deploying or maintaining volitional attention, typically defined by performance deficits on the predictive central arrow cue task, have been subscr ibed to a range of complex disorders such as neglect (e.g., Rafa l , 2000), frontal lobe damage (e.g., Henik, Rafal & Rhodes , 1994), Park inson's d isease (e.g., Kingstone et al , 2002), autism (e.g., 193 Wainwr ight-Sharp & Bryson, 1993; larocci & Burack, 2004), and attention-deficit and hyperactivity disorder (e.g., Pea rson , et al , 1995). The Rist ic and Kingstone (in press) data indicate that these studies and the diagnost ic and rehabilitative appl icat ions that they have spawned (see Park & Ingles, 2001 for a recent review) need to be quest ioned and reevaluated. Finally, a body of attentional literature indicates that reflexive and volitional attention are subserved by distinct neural networks, with reflexive orienting engaging a ventral frontoparietal sys tem and volitional orienting engaging a dorsal frontoparietal sys tem (e.g., Corbetta & Shu lman, 2002). B e c a u s e the vast majority of the investigations that have examined brain networks that subserve reflexive and volitional orienting have used the two c lass ic peripheral and central cuing tasks, the resulting neural distinction that has been proposed may itself be brought into quest ion. Final Ref lect ions and Future Directions The work in this dissertat ion has shown that central spatially nonpredict ive attentional stimuli, such as eyes and arrows, trigger reflexive shifts of attention toward the cued location. Importantly, the reflexive orienting effect triggered by central cues does not s e e m to be equal ac ross all stimuli. G a z e direction produces effects that are strongly reflexive, possibly because the processing of eye direction is subserved by brain mechan isms that appear to be spec ia l ized for that task. Other cues , like arrows, produce behavioral effects that are often similar to those produce by gaze , but in particular situations it has been found that arrows are not as strongly reflexive as gaze cues , e.g. , arrows are more vulnerable to cognitive control and the effects of attentional set than gaze direction (Chapters 5 and 6). Thus, a main proposal to emerge from the present dissertation is that central nonpredict ive attention cues produce reflexive orienting that is best conceptual ized as existing along a cont inuum. 194 Importantly, these findings and ideas depart significantly from the traditional framework that has been appl ied to attentional orienting, where reflexive orienting is triggered by spatially nonpredicit ive peripheral cues and volitional orienting is engaged by spatially predictive central arrow. What is a productive course of action to take when trying to reconci le the present f indings of this thesis within the c lass ic attentional f ramework? One route would be to redefine reflexive and volitional orienting within the c lass ic framework by recogniz ing that the physical location of the attention cues — peripheral versus central — is not a valid indicator of whether a cue triggers reflexive or volitional attention. A s has been d iscussed previously, in very many important ways, both nonpredictive peripheral cues and nonpredict ive central cues produce similar reflexive attention effects. Indeed, the only substantial difference appears to be one cue st imulates parafoveal vis ion and the other st imulates foveal v is ion. What appears to be critical to whether a cue triggers reflexive attention or volitional attention is not where it is presented, but what information it conveys. That is, a reflexive shift in spatial attention is tr iggered by a stimulus cue if that cue does not predict where a target is likely to appear; and a volitional shift in spatial attention occurs if that cue does predict where a target is likely to appear. Here, the distinction would be made based on the reliability of the spatial information of the attention cue and not its spatial posit ion. However, even this redefinition fails to accurately dist inguish between reflexive and volitional orienting. This is because Rist ic and Kingstone (in press) have shown that spatially predictive arrow cues engage both reflexive and volitional p rocesses . Thus it is c lear that the predictive value of the cue alone fails to accurately dist inguish between cues that engage reflexive and volitional attention. At the very least, one would 195 have to include the notion that a predictive attention cue engages volitional attention only if it does not trigger attention reflexively when it is spatially nonpredict ive. A n alternative approach that departs from the c lass ic framework and its notion of reflexive and volitional attention cons iders how an attention cue refers to another object in space (Logan, 1995). What is interesting about this framework, and markedly different from the c lass ic framework, is that the same attentional cue could potentially have very different effects on spatial attention al location depending on the spatial reference frame adopted by an observer. Logan (1995) considered three types of spatial relations. Bas i c spatial relations specify the location of a single object based on the reference frame of the observer. Thus, a basic relation would indicate simply that "The target is there", where the spatial reference frame is centered on the observer. Deictic spatial relations take two arguments and specify the location of one object with respect to the location of another object. This may be expressed in a sentence like "The target is above the cue" . In this case , the location of the target is speci f ied by the reference frame that is centered on the cue. Finally, intrinsic spatial relations specify the location of one object with respect to another object, like deictic relations, except that the spatial reference frame is fixed within the intrinsic axes of the referent object. Thus , according to this conceptual izat ion of attentional cues , it is not a cues ' position in space or even whether a shift is "reflexive" or "volitional" that is important; what is critical is the way the spatial relations between a cue and an object is conveyed. Recent ly, G ibson and Kingstone (in press) extended the original work by Logan (1995) by examining the effects of spatial reference f rames on the orienting effects elicited by peripheral and central cues . Specif ical ly, the authors hypothesized that attentional cues such as peripheral onsets , central arrows and gaze cues , conveyed basic spatial relations between the cue and the target. In contrast, central word cues , 196 such as left, right, above or below, communicated deict ic spatial relations. In keeping with this proposal , G ibson and Kingstone found that peripheral onsets, central arrows and central gaze cues , all produced behavioral effects that were grounded on an observer-centered reference frame; whereas central word cues produced behavioral effects that were grounded on a cue-centered reference frame. What is particularly noteworthy given the findings of the present dissertation is that this alternative framework naturally groups peripheral onsets with central cues like eyes and arrows. In addition it offers a way of gaining insight into differences that may arise between different c l asses of central cues , such as arrows and words. Finally, a more fundamental way of reexamining human attention, and the one that departs most dramatically from the original framework, is to advocate for a research focus that seeks to understand how human attention operates outside the lab and within more complex real world situations. There is a growing body of ev idence indicating that cognit ive and neural p rocesses change as the task situation changes (e.g., Neisser , 1976; Monse l l , 1996). A s a result even the most minor changes within a lab situation can compromise the replicability of an effect (Berry & Klein, 1993; So to -Faraco , Morein-Zamir & Kingstone, 2005; Wolfe & Pokorny, 1990). Col lect ively these studies suggest that attention effects obtained within simple lab settings must be examined against more complex real world sett ings (see Kingstone et al . , 2003). This point is driven home by the present dissertat ion. In trying to address one of the core attention research issues in the field — determining how people use real world attention cues , such as the eyes of others — it was d iscovered that attention is oriented reflexively in response to central nonpredictive cues . In turn, this finding has led to the clear indication for a reevaluation of the c lass ic division of human attention in terms of reflexive orienting that is triggered by peripheral cues and volitional orienting that is triggered by central cues . It is 197 noteworthy that Michael Posne r had long ago anticipated that real world considerat ions may ultimately demand that the c lass ic reflexive/volitional division of human orienting, grounded on the cuing paradigm, may need to be revised substantial ly. At the end of his seminal book Chronometr ic Explorat ions of Mind, he wrote: "I am aware of the speculat ive nature of these relat ionships between the findings of experimental studies and their appl icat ion.. . Whether or not the principles emerging from experimental laboratories and the methods of mental chronometry will be helpful in laying bare some of these mechan isms will remain for the future to determine" (p. 240). 198 References Al l ison, T., Giner, H., McCar thy , G . , Nobre, A . , Puce , A . , Luby, M., & Spencer , D. D. (1994). F a c e recognition in human extrastriate cortex. Journal of Neurophysio logy. 71(2). 821-825. Al l ison, T., P u c e , A . , & McCar thy G . (2000). Soc ia l perception from visual cues . Trends in Cognit ive Sc ience . 4. 267-278. Ba ron -Cohen , S . (1995). Mindbl indness: A n essay on autism and theory of Mind. Massachuse t ts : MIT P ress . Batki, A . , Ba ron -Cohen , S . , Wheelwright, S . , Conne l lan , J . , & Ahluwal ia, J . (2000). Is there an innate module? Ev idence from human neonates. Infant Behavior & Development. 23, 2 2 3 - 2 2 9 . Bentin, S . , Al l ison, T., P u c e , A . , Perez , & McCar thy , G . (1996). Electrophysiological studies of face perception in humans. Journal of Cognit ive Neurosc ience, 8. 551-565. Berry, G . , & Klein, R. (1993). Does motion-induced grouping modulate the f lanker compatibility effect? A failure to replicate Driver & Bavl is . Canad ian Journal of Experimental Psycho logy . 47, 714-729. Br iand, K.A. & Klein, R .M (1987). Is Posner ' s beam the s a m e as Tre isman 's g lue?: O n the relation between visual orienting and feature integration theory. Journal of Experimental Psycho logy : Human Percept ion and Per formance, 13, 2 2 8 - 2 4 1 . C h e a l , M . L. & Lyon, D. R. (1991). Central and peripheral precuing of forced-choice discrimination. Quarterly Journal of Experimental Psycho logy: Human Experimental Psycho logy , 43A, 859-880. Corbetta, M. & Shu lman, G . L. (2002). Control of goal-directed and stimulus-driven attention in the brain. Nature Rev iews Neurosc ience, 3, 201-215. Danziger, S . , & Kingstone, A . (1999). Unmask ing the inhibition of return phenomenon. Percept ion and Psvchophvs ics , 61 . 1024-1037. Driver, J . , Davis , G . , Ricciardel l i , P. , Kidd, P. , Maxwel l , E. , & Baron -Cohen , S . (1999). G a z e perception triggers visuospat ial orienting by adults in a reflexive manner. V isua l Cogni t ion, 6, 509-540. Driver, J . & Spence , C . (1998a). Attention and the crossmodal construction of space . Trends in Cognit ive Sc iences , 2, 254-262. Egly, R., Driver, J . , & Rafa l , R. D. (1994). Shifting visual attention between objects and locations: Ev idence from normal and parietal lesion subjects. Journal of Experimental Psycho logy : Genera l , 123, 161-177. Eimer, M. (1997). Uninformative symbol ic cues may bias visual-spat ial attention: behavioral and electrophysiological ev idence. Biological Psycho logy, 46, 67-71. Enns , J . T., Aus ten , E. L , Di Lollo, V . , Rauschenberger , R. & Yant is , S . (2001). New objects dominate luminance transients in setting attentional priority. Journal of Experimental Psycho logy: Human Percept ion and Per formance, 27, 1283-1302. F a n , J . , McCand l i ss , B. D., Sommer , T., R a z , A . , & Posner , M. , I. (2002). Test ing the Eff iciency and Independence of Attentional Networks. Journal of Cognit ive Neurosc ience, 14, 340-347. Farroni , T., Cs ib ra , G . , S imion, F., & Johnson , M. H. (2002). Eye contact detection in human from birth. Proceedings of the National A c a d e m y of Sc i ences , 99, 9603-9605. F ischer, M.H. , Cas te l , A . D . , Dodd, M.D. , & Pratt, J . (2003). Perceiv ing numbers causes spatial shifts of attention. Nature Neurosc ience, 6, 555-556. Fr iesen, C . K., & Kingstone, A . (1998). The eyes have it!: Ref lexive orienting is triggered by nonpredictive gaze . Psychonomic Bulletin & Review, 5, 490-495. Fr iesen, C . K., & Kingstone, A . (2003). Abrupt onsets and gaze direction trigger independent reflexive attentional effects. Cognit ion, 87, B1 -B10 . Fr iesen, C . K., Rist ic, J . , & Kingstone, A . (2004). Attentional effects of counterpredict ive gaze and arrow cues . Journal of Experimental Psycho logy: Human Percept ion & Per formance. 30, 319-329. Fr ischen, A . & Tipper S . P. (2004). Orienting via observed gaze shift evokes longer term inhibitory effects: Implications for social interaction, attention and memory. Journal of Exper imental Psycho logy: Genera l . 133, 516-533. Gal fano, G . , Ruscon i , E., & Umilta, C . (in press). Number magnitude orients attention, but not against one 's will. Psvchonomic Bulletin & Rev iew. G ibson , B. S . & Bryant, T. A . (2005). Variat ion in cue duration reveals top-down modulation of involuntary orienting to uninformative symbol ic cues . Percept ion & Psvchophvs ics . 67, 749-758. G ibson , B. S . & Kingstone, A . (in press). V isua l attention and the semant ics of space : Beyond central and peripheral cues . Psycholog ica l Sc ience . Hasher, L. & Z a c k s , R. T. (1979). Automat ic and effortful p rocesses in memory. Journal of Exper imental Psycho logy: Genera l . 108, 356-388. Hawkings, H. L , Hil lyard, S . A . , Luck, S . J . , Mou loa , M., Downing, C . J.& Woodward , D. P. (1990). V isua l Attention Modulates Signal Detectability. Journal of Exper imental Psycho logy: Human Percept ion & Per formance, 16. 802-811. Henik, A . , Rafa l , R. D. & Rhodes , D. (1994). Endogenous ly generated and visually guided s a c c a d e s after lesions of human frontal eye fields. Journal of Cognit ive Neurosc ience. 6, 400-412. Hoffman, E. , A . , & Haxby, J . , V . (2000). Distinct representat ions of eye gaze and identity in the distributed human neural sys tem for face perception. Nature Neurosc ience, 3, 80-84. Hommel , B., Pratt, J . , Co lza to , L. & Godi jn, R. (2001). Symbol ic control of visual attention. Psycholog ica l Sc ience , 12, 360-365. Hood, B. M. , Wi l len, J . D., & Driver, J . (1998). Adult 's eye trigger shifts of visual attention in human infants. Psycholog ica l Sc ience , 9, 131-134. Hunt, A . R. & Kingstone, A . (2003a). Inhibition of return: Dissociat ing attentional and oculomotor components . Journal of Exper imental Psycho logy: Human Percept ion and Per formance, 29, 1068-1074. Hunt, A . R. & Kingstone, A . (2003b). Covert and overt voluntary attention: Linked or independent? Cognit ive Brain R e s e a r c h , 18, 102-105. larocci, G . & Burack, J . A . (2004). Intact covert orienting to peripheral cues among children with aut ism. Journal of Aut ism & Developmental Disorders, 257-264. Jon ides , J . (1980). Toward a model of the mind's eye 's movement. Canad ian Journal of Psycho logy. 34, 103-112. Jon ides , J . (1981): Voluntary versus automatic control over the mind's eye's movement. In J . B. Long and A . D. Baddeley (Eds.) . Attention & Per formance IX (pp. 187-203). Hi l lsdale, N J : Er lbaum. Jon ides , J . , Naveh-Benjamin, M. , & Palmer, J . (1985). A s s e s s i n g automaticity. Ac ta Psvcho log ica , 60, 157-171. Kahneman , D. & Tre isman, A . (1984). Chang ing v iews of attention and automaticity. In R. Parasuraman & D. A . Davies (Eds.) , Variet ies of Attention, (pp. 29-61). S a n Diego, C A : Academ ic P ress . Kanwisher , N. (2000). Domain-specif ic i ty in face perception. Nature Neurosc ience, 3, 759-763. Kanwisher , N., McDermott , J , & Chun , M. , M. (1997). The fusiform face area: A module in human extrastriate cortex spec ia l ized for face perception. The Journal of Neurosc ience. 17, 4302-4311. Kingstone, A . , Klein, R . M . , Morein-Zamir, S . , Hunt, A . , Maxner , C , & Fisk, J . (2002). Components of attention: Chronometr ic techniques and Park inson 's d isease. Journal of Cl in ical and Exper imental Neuropsychology, 24, 951-967. Kingstone, A . , Smi lek, D., Ristic, J . , Fr iesen, C . K., & Eas twood, J . D. (2003). Attention researchers! It is time to look at the real world. Current Direct ions in Cognit ive Sc ience , 12, 176-184. Kingstone, A , Tipper, C , Rist ic, J & Ngan, E (2004). The eyes have it!: A n fMRI investigation. Brain and Cogni t ion, 55, 269-271. Klein, R . M . (1988). Inhibitory tagging system facilitates visual search . Nature, 334, 4 3 0 -431 Klein, R. M., Kingstone, A . & Pontefract, A . (1992). Orienting of visual attention. In K. Rayner (Ed.), Eve Movements and V isua l Cognit ion: S c e n e Percept ion and Read ing, (pp. 46-63). North-Hol land: Elsevier Sc ience Publ ishers B. V . Klein, R. M. (2004). O n the control of v isual orienting. In M.I. Posner , (Ed.) Cognit ive Neurosc ience of Attention. New York: Guilford P ress . Kobayash i H, & Kosh ima S . (1997). Unique morphology of the human eye. Nature. 387, 767-768. Langton, S . R. H., & Bruce, V . (1999). Ref lexive socia l orienting. V isua l Cogni t ion. 6, 541-567. Langton, S , R, H. & Bruce, V . (2000). Y o u must see the point: Automat ic processing of cues to the direction of socia l attention. Journal of Exper imental Psycho logy: Human Percept ion & Per formance, 26, 747-757. Logan, G . D. (1995). Linguistic and conceptual control of visual spatial attention. Cognit ive Psycho logy , 28, 103-174. Luck, S . J . , Hil lyard, S . A . , Mouloa, M., Woldorff, M. G . , Clark, V . P. & Hawkings, H.L. (1994). Effects of spatial cuing on luminance detectability: Psychophys ica l and Electrophysiological ev idence for early select ion. Journal of Exper imental Psycho logy: Human Percept ion & Per formance, 887-904. Luck, S . J . , W o o d m a n , G . F. & Voge l , E. K. (2000). Event-related potential studies of attention. Trends in Cognit ive S c i e n c e s , 4, 432-440. Maruff, P., Yuce l , M. , Danckert, J . , Stuart, G . & Curr ie, J . (1999). Facil itation and inhibition arising from the exogenous orienting of covert attention depends on the temporal properties of spatial cues and targets. Neuropsycholog ia , 37, 731-744. McDona ld , J . , J . , Teder-Sale jarv i , W . A . , Heraldez, D., & Hil lyard, S . A . (2001). Electrophysiological ev idence for the "missing link" in c rossmoda l attention. Canad ian Journal of Exper imental Psycho logy, 55, 141-149. McDona ld , J . , J . & Ward , L., M. (2000). Involuntary listening aids see ing : Ev idence from human electrophysiology. Psycholog ica l Sc ience , 11, 167-171. Monse l l , S . (1996). Control of mental p rocesses . In V . Bruce (Ed.), Unso lved Myster ies of the Mind: Tutorial E s s a y s in Cognit ion (pp. 93-148). Howe, S u s s e x : Er lbaum (UK) Taylor & Francis . Moscov i tch , M. , Winocur, G . , & Behrmann, M. (1997). What is specia l about face recognit ion? Nineteen exper iments on a person with visual object agnos ia and 204 dyslexia but normal face recognit ion. Journal of Cognit ive Neurosc ience, 9, 555-604. Muller, H, J . & Findlay, J , M. (1988). The effect of visual attention on peripheral discrimination thresholds in single and multiple element d isplays. Ac ta Psvcho log ica . 69, 129-155. Muller, H. J . , & Rabbitt, P. M. A . (1989). Reflexive and voluntary orienting of visual attention: time course of activation and resistance to interruption. Journal of Experimental Psycho logy : Human Percept ion and Per formance, 15, 315-330. Muller, H. J . & Humphreys, (1991). Luminance increment detect ion: Capac i ty - limited or not? Journal of Exper imental Psycho logy: Human Percept ion & Per formance, 17, 107-124. Naveh-Ben jamin , M. & Jon ides , J . (1986). O n the automaticity of f reguency coding: Effects of compet ing task load, encoding strategy and intention. Journal of Experimental Psycho logy : Human Percept ion and Per formance, 12, 378-386. Neisser , U. (1976). Cognit ion and Reali ty. S a n Franc isco: F reeman . Palmer, J . & Jon ides , J . (1988). Automat ic memory search and the effects of information load and irrelevant information. Journal of Exper imental Psycho logy : Human Percept ion and Per formance. 14, 136-144. Park, N. W . & Ingles, J . L. (2001). Effect iveness of attention rehabilitation after an acquired brain injury: A meta-analysis. Neuropsychology, 199-210. Pearson , D. A . , Yaf fee, L .S . , Love land, K. A . , & Norton, A . M. (1995). Covert visual attention in chi ldren with attention deficit hyperactivity disorder: Ev idence for developmental immaturity? Development & Psvchopatho logy, 7, 351-367. Perrett, D., I. (1985). V isua l cel ls in the temporal cortex sensit ive to face view and gaze direction. Proceed ings of the Roya l Society of London-Ser ies B: Biological Sc iences , 223. 293-317. Posner , M. I. (1980). Orienting of attention. Quarterly Journal of Exper imental Psycho logy, 32, 3-25. Posner , M. (1978). Chronometr ic Explorat ions of the Mind. Hi l lsdale, N J : Er lbaum. Posner , M. I, Snyder , C . R. R, & Dav idson, B. J . (1980). Attention and detection of s ignals. Journal of Exper imental Psycho logy: Genera l , 109, 160-174. Posner , M. I., & C o h e n , Y . (1984). Components of V isua l Orienting. In H. B o u m a & D. G . Bowhuis (Eds.) , Attention & Per formance X (pp. 531-556). Hi l lsdale, N J : Er lbaum. Pratt, J . & Ab rams , R.A. (1995) Inhibition of return to success ive ly cued spatial locations. Journal of Exper imental Psycho logy: Human Percept ion and Per formance, 21 , 1343-1353. Pratt, J . & Hommel , B. (2003). Symbol ic control of v isual attention: The role of working memory and attentional control settings. Journal of Exper imental Psycho logy : Human Percept ion & Per formance, 29, 835-845. Pr ime, D. J . & Ward , L. M. (2004). Inhibition of return from stimulus to response. Psycholog ica l Sc ience . 15. 272-275. Rafal , R.D. , Ca labres i , P.A. , Brennan, C W . & Sciolto, T .K. (1989). S a c c a d e preparation inhibits reorienting to recently attended locations. Journal of Exper imental Psycho logy: Human Percept ion & Per formance, 15, 6 7 3 - 6 8 5 . Rafa l , R. D. (2000). Neglect II: Cognit ive neuropsychological issues. In M. J . Farah & T. E. Feinberg (Eds.) , Patient B a s e d Approached to Cognit ive Neurosc ience (pp. 125-141). Cambr idge , M A : MIT P ress . Reuter-Lorenz, P. A . , J h a , A . & Rosenquis t , J . N. (1996). What is inhibited in inhibition of return? Journal of Exper imental Psycho logy: Human Percept ion and Per formance, 22, 367-378. Rist ic, J . , Fr iesen, C . K., & Kingstone, A . (2002). A re eyes spec ia l? It depends on how you look at it. Psvchonomic Bulletin & Rev iew, 9, 507-513. Rist ic, J . , & Kingstone, A . (2005). Taking control of reflexive social attention. Cognit ion, 94, B55 -B65 . Rist ic, J . & Kingstone, A . (in press). Attention to arrows: Pointing to a new future. Quarterly Journal of Experimental Psycho logy. Rist ic, J . , Wright, A . & Kingstone, A . (in press). The number line effect reflects top-down control. Psvchonomic Bulletin & Rev iew. Schul ler , A . , & Ross ion , B. (2001). Spat ial attention triggered by eye gaze increases and speeds up early visual activity. Neuroreport,12, 2381-2386. Schul ler , A . , & Ross ion , B. (2004). Percept ion of static eye gaze direction facilitates subsequent early visual process ing. Cl in ical Neurophysio logy, 115, 1161-1168. Snyder , J . & Kingstone, A . (In press). Inhibition of return at multiple locations and its impact on visual search . V isua l Cognit ion. S p e n c e , C . & Driver, J . (1994). Covert spatial orienting in audit ion. Journal of Experimental Psycho logy : Human Percept ion and Per formance, 20, 555-574. Spence , C . & Driver, J . (1997). Audiov isual links in exogenous covert spatial orienting. Percept ion & Psychophvs ics , 59, 1-22. Spence , C , Nichhol ls, M . E. R., Gi l lespie, N & Driver, J . (1998). C ross -moda l links in exogenous covert spatial orienting between touch, audition and vis ion. Psvchonomic Bulletin & Review. 60, 544-557. Soto-Faraco , S . , Morein-Zamir, S . & Kingstone, A . (2005). O n audiovisual spatial synergy: The fragility 'of the phenomenon. Percept ion & Psvchophvs ics , 67, 444-457. So to -Faraco , S . , Sinnett, S . , A ls ius , A . & Kingstone, A . (2005). Spat ia l orienting of tactile attention induced by gaze direction. Psychonomic Bulletin & Rev iew, 12, 1024-1031. Symons , L. A . , Lee, K, Cedrone, C . C , Nishimura, M. (2004). What are you looking at? Acuity for triadic eye gaze . Journal of Genera l Psycho logy, 13, 451-469. Taylor, T. L., & Klein, R. M. (1998). On the causes and effects of inhibition of return. Psychonomic Bulletin and Rev iew, 5, 6 2 5 - 6 4 3 . Theeuwes , J . (1991). Exogenous and endogenous control of attention: The effects of visual onsets and offsets. Percept ion & Psvchophvs ics , 49, 83-90. Theeuwes , J . , Kramer, A . F., Hahn, S . , & Irwin, D. E. (1998). Our eyes do not a lways go where we want them to go: capture of the eyes by new objects, Psycholog ica l Sc ience , 9, 379 -385. Tipples, J . (2002). Eye gaze is not unique: Automat ic orienting in response to noninformative arrows. Psychonomic Bulletin & Rev iew, 9, 314-318. Todd, J . T. & van Gelder , P. (1979). Implications of a transient-sustained dichotomy for the measurement of human performance. Journal of Exper imental Psycho logy : Human Percept ion and Per formance, 5, 625-638. Vui l leumier, P. et al (2001). Neural fate of seen and unseen faces in visuospat ial neglect: A combined event-related functional MRI and event-related potential study. Proceed ings of the National A c a d e m y of Sc ience , U S A , 98, 3495-3500. Wainwright-Sharp, J . A . , & Bryson, S . E . (1993). V isua l orienting deficits in high functioning people with aut ism. Journal of Aut ism and Developmenta l Disorders. 23, 1-13. Ward , L. M. (1994). Supramoda l and modali ty-specif ic mechan isms for st imulus-driven shifts of auditory and visual attention. Canad ian Journal of Exper imental Psycho logy. 48, 242-259. Ward , L. M. , McDona ld , J . J . , & Lin, D. (2000). O n asymmetr ies in c rossmodal spatial attention orienting. Percept ion & Psychophys ics . 62, 1258-1264. Warner, C . B., Juo la , J . F. & Koshino, H. (1990). Voluntary al location versus automatic capture of v isual attention. Percept ion & Psychophys ics . 48, 243-251. Wol fe, J . M., & Pokorny, C . W. (1990). Inhibitory tagging in visual search : A failure to replicate. Percept ion & Psychophys ics . 48, 357-362. Yant is , S . & Jon ides , J . (1990). Abrupt visual onsets and select ive attention: Voluntary vs . automatic al location. Journal of Exper imental Psycho logy: Human Percept ion and Per formance. 16, 121-134. Yant is , S . , & Hil lstrom, A . P. (1994). St imulus-dr iven attentional capture: Ev idence from equiluminant visual objects. Journal of Exper imental Psycho logy: Human Percept ion and Per formance. 20, 95-107. Yant is , S . , & Egeth, H. E. (1999). O n the distinction between visual sa l ience and st imulus-driven attentional capture. Journal of Exper imental Psycho logy: Human Percept ion and Per formance. 25, 661-676. A P P E N D I X I U B C Behaviora l Resea rch Ethics Board Certif icate of Approva l 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0100574/manifest

Comment

Related Items