The role of attention in Simon effect asymmetries

The Simon effect (Craft & Simon, 1970; Simon & Rudell, 1967) is a phenomenon due to the automatic coding of the spatial position of the stimulus, which, in turn, activates the corresponding response, thus causing interference at the response selection stage (e.g., Kornblum, Hasbroucq & Osman, 1990; Zorzi & Umiltà, 1995): When participants have to select the left or right response on the basis of a non-spatial feature of the stimulus, reaction times (RTs) are faster if the position of stimulus and response correspond than if they do not. Tagliabue, Vidotto, Umiltà, Altoè, Treccani and Spera (2007) showed that this effect is often asymmetric, being greater in the right than in the left side of space in strong right-handers and greater in the left than in the right side of space in strong left-handers. They distinguished between the stimulus Simon effect (stSE), that is the Simon effect measured by comparing the two responses to the same stimulus and the response Simon effect (reSE), that is the Simon effect measured by comparing the same response to the two stimuli. Therefore, there can be two types of asymmetries, which originate by comparing the right stSE with the left stSE (i.e., the stSE asymmetry) and by comparing the right reSE with the left reSE (i.e., the reSE asymmetry). They proposed that these asymmetries were caused by cerebral lateralization of the two mechanisms involved in the Simon effect, respectively attention orienting and response selection. In this work we focused mostly on the stSE asymmetry, by hypothesizing that this would depend on the lateralization of attentional mechanisms. The role of attentional processes in the Simon effect is supported by the attentional shift hypothesis (Nicoletti & Umiltà, 1994; Stoffer, 1991; Stoffer & Umiltà, 1997); moreover, we know that attention basically depends on mechanisms predominantly localized, in right-handers, in the right hemisphere. Mesulam’s model (Mesulam, 1981; 1999) proposed that the right hemisphere would control attention shifts toward the right and the left hemi-space, whereas the left hemisphere would control attention shifts only toward the right hemi-space, and it is supported by several evidences. Evidences to this model come from both disease (Heilman, Watson & Valenstein, 2003; Mesulam, 1981; Mesulam, 1999; Posner, Walker, Friedrich & Rafal, 1984; Rafal, 1999) and healthy states (Corbetta, Miezin, Shulman & Petersen, 1993; Kim, Gitelman, Nobre, Parrish, LaBar & Mesulam, 1999; Nobre, Sebestyen, Gitelman, Mesulam, Frackowiak & Frith, 1997). Our hypothesis is that, because of this lateralization, in a Simon task, more attention is allocated on the stimulus when it appears on the right than on the left because of the reasons due to the fact that both hemispheres orient attention on the right stimulus and only one orients the attention on the left stimulus, and to confirm our hypothesis we used a Posner paradigm (Posner, Snyder & Davidson, 1980). If it is true that an event on the right catches more attention than the same event on the left, in this task the cue should be more salient when it appears on the right, and therefore we would expect to observe more benefits (RT faster in valid trials) and more costs (RT slower in invalid trials) when the cue appears in this position than when it appears on the left. Some asymmetries in this paradigm were already observed by Downing and Pinker (1985) and by Gawryszewski, Riggio, Rizzolati and Umiltà (1987), who reported a difference in the attentive effects in the two visual hemi-fields (VHFs). The first part of the present research is composed by four experiments, all involving a peripheral cue. To investigate both exogenous and endogenous components of attention (Jonides, 1981; Umiltà, 2000), in Experiments 1 and 3 the cue was informative (which triggers both exogenous and endogenous attentional shifts) and in Experiments 2 and 4 it was non-informative (which triggers purely exogenous attentional shifts). Each participant performed one of the four versions of the Posner task and, one week later, a Simon task. In Experiment 1 and Experiment 2 we used short SOAs (Stimulus Onset Asynchrony), and in Experiment 3 and Experiment 4 we used long SOAs, in order to investigate if the asymmetries were present in both effects usually observed when the exogenous mechanism is involved: respectively, the facilitation and the Inhibition of Return (IoR) (Posner & Cohen, 1984; Posner, Rafal, Choate & Vaughan, 1985). We decided to test both right- and left-handed participants, because of the different lateralization of attentional mechanisms in these two populations (Bradshaw, 1989; Bryden, 1982; DeLuca, 1993; Dronkers & Knight, 1988; Pantano, Frontoni, Iacoboni, Di Piero & Lenzi, 1992; Wang, 1980). Tagliabue et al. (2007) reported a greater Simon effect on the right in strong right-handers and a greater Simon effect on the left in strong left-handers. Participants with less handedness lateralisation (with a score of at least 70% instead of 90% with the Edinburgh Handedness Inventory; Oldfield, 1971) were tested, and it was hypothesized that asymmetries in the same direction of Tagliabue et al. (2007) in right-handers and either in the same direction or not present in left-handers would be found. Our main hypotheses were two: The first one was that participants with different handedness would show asymmetries in the opposite direction in both tasks; the second one was that, regardless of handedness, participants with a strong asymmetry in one direction in the Posner task would show the asymmetry in the same direction in the Simon task. The work is essentially divided in three parts: In the first and second part, the first and the second hypothesis were investigated, respectively. In the third part some control conditions were added. Experiment 1 and Experiment 2 demonstrated that the facilitation effect was greater on the right than on the left in right-handers and greater on the left than on the right in left-handers, and confirmed our hypothesis about a difference in benefits and costs in the two VHF depending on handedness. Even though the facilitatory effect was greater when the cue was informative than when it was non informative – because of the involvement of the endogenous component of attention – this component does not interact with the asymmetries, that are indeed present in both cases. Concerning the Simon task, data of both experiments shows a symmetric Simon effect in left-handers and a Simon effect greater on the right in right-handers. The results of Experiment 3 and Experiment 4 differ from the previous experiments, by depending on the informativeness of the cue. Indeed, it is only when an informative cue was employed (Experiment 3) that we observed the asymmetries, and these were present only in right-handers. Left-handers did not show IoR, and right-handers shown IoR only when the cue appears on the left. Interestingly, a global IoR is not present, and the absence of this effect when an informative cue is employed is a well-established fact (e.g. Umiltà, 2000), because of the involvement of the endogenous component of attention. However, this component seems to have a different effectiveness in the two VHFs. Indeed, when participants are asked to maintain attention on the right, the endogenous component efficiently opposes the automatic apparition of the IoR (that is indeed absent). However, when participants are asked to maintain attention on the left, this component is not effective enough to avoid the IoR (that is indeed present). Concerning the Simon task, data of both experiments showed a symmetric Simon effect in left-handers and a Simon effect greater on the right in right-handers. Globally, the results of the four Posner tasks showed that the facilitation effect was greater when the cue appears on the right in right-handers and when the cue appears on the left in left-handers. The facilitation effect was greater when the cue was informative than when it was non informative, because of the involvement of the endogenous component of attention, but this component did not interact with the asymmetries that were present in both cases. On the contrary, it did interact with the asymmetries when the SOA increases. The advantage of the right VHF in right-handers was evident only when an endogenously driven shift of attention took place (i.e., only in Experiment 3). Data suggest that attention orienting, both exogenous and endogenous, is advantaged when it is toward the right rather than the left, because in the former case both hemispheres participate, while in the latter case only one participates. In other words, data seem to suggest that when the cue appears on the right, movement and maintenance of attention are more efficient. Concerning the left-handers, data showed the opposite trend (i.e., an advantage of the left VHF). The results of the four Simon tasks, instead, showed an effect greater on the right in right-handers and symmetric in left-handers. Therefore, the first part of this work confirms our first hypothesis, which was that participants with different handedness will show different patterns of asymmetries, in both Posner and Simon tasks. The purpose of the second part of the work was to investigate the second hypothesis, that regardless of handedness, participants with strong asymmetries in one direction in the Posner task should show strong asymmetries in the same direction in the Simon task. For this reason we analyzed together the Simon data of all participants. First of all, an overall analysis on the performance in the Simon task shown that the Simon effect is greater on the right in right-handers and symmetric in left-handers. The fact that the effect is symmetric, instead of greater on the left, in left-handers, indicates that the direction of the asymmetries is not due only to the preference of the dominant hand. Subsequently, we split all the participants into two groups, based on the direction of the asymmetries shown in the Posner task: One group with a strong advantage on the right VHF and one group with a strong advantage of the left VHF, and we analyzed the performance of the two groups in the Simon task. Handedness and VHF advantage interfered with the stSE: When preffered hand and advantaged VHF correspond (i.e., in right-handers with right VHF advantage and left-handers with left VHF advantage), the stSE was greater in the spatial position that was advantaged by both mechanisms, respectively, right and left.. When preffered hand and advantaged VHF did not correspond (i.e., in right-handers with left VHF advantage and left-handers with right VHF advantage) the left stSE disappeared. Data of the two groups support the hypothesis that Simon effect asymmetries are due to the cooperation of the two mechanisms involved in the effect, respectively, attention and selection of the response. In the third part of the work we conducted two control experiments to exclude the possibility that results observed in the first part of the work were not due to the interference of a confounding variable (i.e., the dominant hand used by participants to perform the Posner task and the order of the tasks). Therefore, in Experiment 5 we tested two new groups of 30 participants each. Participants had to respond with the right hand in half of the blocks and with the left hand in the other half. Moreover, half of participants performed the Posner task first and the Simon task one week later (as in the previous experiments), whereas the other half performed the Simon task first and the Posner task one week later. In Experiment 6 participants performed only a Posner task and they had to give a bimanual response (i.e., by pressing simultaneously two keys). The results of Experiment 5 (the procedure was identical to that of Experiment 1, except for the hand used) showed that the presence of asymmetries depends on the hand the participants use to respond. Asymmetries were observable only when the participants perform the task with their right (dominant) hand. This confirms that the hand the participants use to respond can affect the asymmetries observed in previous experiments. However, this result provides evidence the results obtained in previous experiments are not due only to the hand used to perform the task, because in this case we would expect an asymmetry in the opposite direction (greater facilitation effect on the left) when the participants use their left (non dominant) hand. Therefore, the lack of any asymmetry with the left hand could be explained with an automatic shift of attention toward the effector (see Eimer, Forster, Van Velzen & Prabhu, 2005; Forster & Eimer, 2007). In Experiment 6 participants had to give a simultaneous bimanual response. In this case we used both a short and long SOAs to analyze both facilitation and IoR. Analysis conducted on RTs of the first key pressed by the participants showed that the asymmetries are globally present only when short SOA are employed and that, again, they are going in the same direction of the previous experiments. In conclusion, this work demonstrates that handedness could not, alone, give an exhaustive explanation of the Simon effect asymmetries. Obviously, it is not easy to analyze separately the two mechanisms involved in this task – attention orienting and response selection – and this is confirmed by the two control experiments, which show that they are strongly related in a simple RT task like a Posner task too. However, on the basis of this work, we can conclude that the direction of the attentive asymmetries, and not only the handedness, explains the Simon effect asymmetries.L'effetto Simon (Craft & Simon, 1970; Simon & Rudell, 1967) è un fenomeno causato dalla codifica automatica della posizione dello stimolo (Kornblum, Hasbroucq & Osman, 1990; Zorzi & Umiltà, 1995): Quando i partecipanti devono selezionare una risposta sulla base di una caratteristica non spaziale dello stimolo, i TR (tempi di reazione) sono più veloci quando la posizione dello stimolo e della risposta corrispondono. Tagliabue, Vidotto, Umiltà, Altoè, Treccani e Spera (2007) hanno dimostrato che questo effetto è asimmetrico, essendo maggiore a destra nei destrimani e maggiore a sinistra nei mancini, e hanno distinto tra stimulus Simon effect (stSE) cioè l’effetto Simon misurato confrontando le due risposte allo stesso stimolo, e response Simon effect (reSE) cioè l’effetto Simon misurato confrontando la stessa risposta ai due stimoli. Quindi, ci sono due tipi di asimmetrie, che si ottengono confrontando lo stSE di destra e di sinistra (l’asimmetria dello stSE) e confrontando il reSE di destra e di sinistra (l’asimmetria del reSE), e gli autori hanno proposto che queste asimmetrie siano causate dalla lateralizzazione dei meccanismi che sono coinvolti nell’effetto Simon, rispettivamente l’orientamento dell’attenzione e la selezione della risposta. In questo lavoro ci siamo concentrati principalmente sull’asimmetria dello stSE, e abbiamo indagato se effettivamente questa fosse causata dalla lateralizzazione dei meccanismi attentivi. Il ruolo dell’attenzione nell’effetto Simon è confermato dalla teoria dell’attentional shift (Nicoletti & Umiltà, 1994; Stoffer, 1991; Stoffer & Umiltà, 1997); inoltre sappiamo che l’attenzione dipende da meccanismi principalmente localizzati, nei destrimani, nell’emisfero di destra. Il modello di Mesulam (Mesulam, 1981; 1999) propone che l’emisfero di destra orienti l’attenzione verso entrambi gli emi-campi visivi, e che invece l’emisfero di sinistra orienti l’attenzione nel solo emi-campo visivo di destra, ed è confermato da numerose evidenze, provenienti sia da ricerche su pazienti con lesioni cerebrali (Heilman, Watson e Valenstein, 2003; Mesulam, 1981; Mesulam, 1999; Posner, Walker, Friedrich e Rafal, 1984; Rafal, 1999) sia da ricerche su soggetti normali condotti attraverso l’uso di tecniche di neuro-immagine (Corbetta, Miezin, Shulman e Petersen, 1993; Kim, Gitelman, Nobre, Parrish, LaBar e Mesulam, 1999; Nobre, Sebestyen, Gitelman, Mesulam, Frackowiak e Frith, 1997). L’idea di partenza di questo lavoro è che, a causa di questa lateralizzazione, in un compito Simon lo stimolo a destra riceva più attenzione dello stimolo a sinistra perché nel primo caso si attivano entrambi gli emisferi mentre nel secondo se ne attiva uno solo, e per confermare questa ipotesi è stato utilizzato un paradigma di tipo Posner (Posner, Snyder & Davidson, 1980). Se è vero che un evento a destra riceve più attenzione dello stesso evento a sinistra, in questo tipo di paradigma il cue dovrebbe essere più saliente quando compare a destra, e quindi si dovrebbero osservare più benefici (TR più veloci nelle prove valide) e più costi (TR più lenti nelle prove invalide) quando il cue compare in questa posizione piuttosto che a sinistra. Alcune asimmetrie in questo tipo di paradigma erano già state osservate da Downing e Pinker (1985) e da Gawryszewski, Riggio, Rizzolati e Umiltà (1987), i quali riportano una differenza nei costi attentivi a seconda dell’emi-campo visivo in cui compare il cue. La prima parte della ricerca è composta da quattro esperimenti, tutti condotti utilizzando un cue periferico. Allo scopo di indagare sia la componente esogena che quella endogena dell’attenzione (Jonides, 1981; Umiltà, 2000), negli Esperimenti 1 e 3 il cue era informativo (questo tipo di cue elicita entrambe le componenti), e negli Esperimenti 2 e 4 il cue era non informativo (questo tipo di cue elicita solamente la componente esogena). Ogni partecipante veniva assegnato casualmente ad uno dei quattro esperimenti e successivamente partecipava anche ad un compito Simon. Sono state utilizzate SOA (Stimulus Onset Asynchrony) brevi negli Esperimenti 1 e 2 e SOA lunghe negli Esperimenti 3 e 4, allo scopo di investigare se le asimmetrie fossero presenti in entrambi gli effetti solitamente osservati quando il meccanismo esogeno viene indagato: La facilitazione e l’Inibizione di Ritorno (IoR) (Posner & Cohen, 1984; Posner, Rafal, Choate e Vaughan, 1985). Sono stati testati, per ogni esperimento, un gruppo di 30 destrimani e uno di 30 mancini, in quanto sappiamo che queste due popolazioni hanno una diversa lateralizzazione cerebrale dei meccanismi attentivi (Bradshaw, 1989; Bryden, 1982; DeLuca, 1993; Dronkers & Knight, 1988; Padovani, Pantano, Frontoni, Iacoboni, Di Piero e Lenzi, 1992; Wang, 1980). Tagliabue et al. (2007) riportano un effetto Simon maggiore a destra in partecipanti con una forte preferenza della mano destra e maggiore a sinistra in partecipanti con una forte preferenza della mano sinistra. Noi abbiamo testato partecipanti con una preferenza manuale meno marcata – con un punteggio minimo di 70% invece di 90% all’Edinburgh Handedness Inventory (Oldfield, 1971) – ipotizzando di trovare un’asimmetria nella stessa direzione di quella osservata da Tagliabue et al. (2007) nei destrimani e nella stessa direzione o non presente nei mancini. Le ipotesi su cui ci siamo concentrati sono due: La prima, che gruppi di partecipanti con una diversa preferenza manuale mostrino asimmetrie in direzioni diverse in entrambi i compiti; la seconda, che, a prescindere dalla preferenza manuale, i partecipanti con una forte asimmetria in una direzione nel compito Posner mostrino la stessa asimmetria nella stessa direzione nel compito Simon. Il lavoro è sostanzialmente diviso in tre parti: Nella prima è stata indagata la prima ipotesi, nella seconda la seconda ipotesi, e nella terza sono state aggiunte alcune condizioni di controllo. Gli Esperimenti 1 e 2 mostrano che l’effetto di facilitazione è maggiore a destra nei destrimani e maggiore a sinistra nei mancini, confermando che i costi e i benefici nei due emi-campi visivi sono diversi in funzione della preferenza manuale. Anche se l’effetto di facilitazione è maggiore quando il cue è informativo piuttosto che non informativo a causa dell’intervento della componente endogena dell’attenzione, questa componente non interagisce con le asimmetrie, che sono presenti in entrambi gli esperimenti. Per quanto riguarda il compito Simon, entrambi gli esperimenti mostrano un effetto Simon simmetrico nei mancini e maggiore a destra nei destrimani. Gli esperimenti 3 e 4 invece, diversamente dai precedenti, mostrano andamenti diversi a seconda dell’informatività del cue. Infatti, è solo quando viene utilizzato il cue informativo che si possono osservare le asimmetrie, e solo nei destrimani. I mancini non mostrano IoR, mentre questo effetto è presente nei destrimani quando il cue compare a sinistra. È interessante notare che, globalmente, in accordo con la letteratura sull’argomento (Umiltà, 2000), la IoR non è presente, a causa dell’intervento della componente endogena dell’attenzione. Questa tuttavia sembra avere una diversa efficacia nei due emi-campi visivi: Quando al partecipante è richiesto di mantenere l’attenzione a destra la componente endogena contrasta efficacemente l’instaurarsi automatico della IoR (che, infatti, non è presente), mentre quando al partecipante è richiesto di mantenere l’attenzione a sinistra, questa componente non è altrettanto efficace (infatti l’IoR è presente). Per quanto riguarda il compito Simon, entrambi gli esperimenti mostrano un effetto Simon simmetrico nei mancini e maggiore a destra nei destrimani. Complessivamente, i risultati dei quattro compiti Posner mostrano che la facilitazione è maggiore quando il cue compare a destra nei destrimani e quando il cue compare a sinistra nei mancini. L’effetto è globalmente maggiore quando il cue è informativo rispetto a quando non lo è, e questo è dovuto all’intervento della componente endogena, che però non interagisce con le asimmetrie, che s