Journal of

Journal of Memory and Language 53 (2005) 359–377

www.elsevier.com/locate/jml

Memory andLanguage

Phonological activation of ignored pictures: Furtherevidence for a cascade model of lexical accessq

Eduardo Navarrete, Albert Costa *

GRNC, Parc Cientıfic Universitat de Barcelona & Hospital Sant Joan de Deu, Departament de Psicologia Basica,

Universitat de Barcelona, Spain

Received 18 March 2005; revision received 19 May 2005Available online 7 July 2005

Abstract

Four experiments are reported exploring whether distractor pictures activate their phonological properties in thecourse of speech production. In Experiment 1, participants were presented with two pictures and were asked to nameone while ignoring the other. Distractor pictures were phonologically related, semantically related or unrelated to thetarget picture. Naming latencies were faster in the phonologically related condition (the phonological facilitationeffect—PFE) than in the unrelated condition. No difference between semantically related and unrelated distractorswas observed. In Experiment 2, participants were asked to name the color in which a picture was presented while ignoringthe depicted object. Naming latencies were faster when colors and objects were phonologically related. In Experiments 3and 4, the PFE was replicated under slightly different experimental conditions. Together, these results reveal that distrac-tor pictures that are irrelevant to the communicative intention of the speaker activate their phonological content. Thisobservation supports the notion that activation flows in a cascaded manner through the speech production system.� 2005 Elsevier Inc. All rights reserved.

Keywords: Speech production; Cascade processing; Ignored picture processing

Introduction

Speech production involves the retrieval of at leastfour different types of information: conceptual, lexical,

0749-596X/$ - see front matter � 2005 Elsevier Inc. All rights reserv

doi:10.1016/j.jml.2005.05.001

q This research was supported by grants from Ministerio deCiencia y Tecnologıa (BSO2002-01545) and the James S.McDonnell Foundation (JSMF-20002079). Albert Costa wassupported by the ‘‘Ramon y Cajal program’’ from the Minis-terio de Ciencia y Tecnologıa. We thank Bradford Mahon,Nuria Sebastian-Galles, Salvador Soto-Faraco, Scott Sinnett,Mikel Santesteban, Iva Ivanova, and Gavin Burgess, for theircomments on the manuscript.* Corresponding author. Present address: Departament de

Psicologia Basica, Universitat de Barcelona, Pg. Vall d�Hebron171, 08035 Barcelona, Spain.

E-mail address: acosta@ub.edu (A. Costa).

phonological, and articulatory. For example, in thecourse of naming a picture, speakers need to retrievethe conceptual representation corresponding to the pic-ture, select the respective lexical representation, retrieveits phonological content, and finally retrieve and pro-duce the articulatory gestures. One of the issues thatmodels of speech production need to address consistsof what information is active at each of these levels inthe course of speech production.

It is generally assumed that when accessing the targetconceptual representation (e.g., dog) other semanticallyrelated conceptual representations (e.g., cat, horse) be-come activated as well (e.g., Caramazza, 1997; Dell,1986; Levelt, 1989). In this scenario, the question iswhether conceptual activation percolates to the lexicalsystem. That is, would any activated conceptual repre-

ed.

Fig. 1. Schematic representation of the flow of activation in three different models. The arrows represent flow of activation and thecircles the conceptual, lexical, and sublexical representations. The thickness of arrows and circles represents the magnitude of theactivation. The cascade model, the discrete model proposed by Levelt et al. (1999) and the model proposed by Bloem and La Heij(2003) are described in panels A, B, and C, respectively.

1 This model is silent on whether or not activation flows in acascade fashion from the lexical to the phonological level.However, given that it is a modification of Starreveld and LaHeij�s model (1995, 1996), it presumably maintains the assump-tion of cascade processing between these two levels ofrepresentation.

360 E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377

sentation spread some activation to its lexical represen-tation during the course of lexicalization? And, if so,would activated lexical representations spread some acti-vation to the phonological level? These are importantquestions because the types of processes in charge ofencoding/selecting information at each of these levelsmay differ depending on what is activated at each level.In this article, we investigate these issues by exploringthe extent to which stimuli that are not relevant forthe speaker�s communicative goal activate their corre-sponding phonological information during picture nam-ing. Specifically, we explore whether there isphonological activation of ignored picture names inthe course of speech production.

There are at least three different proposals regardinghow information is passed in a feed-forward fashionfrom level to level of representation (see Fig. 1). Full-cascade models assume that any activated representa-tion at a given processing level spreads a proportion ofits activation to its immediately linked representationat the subsequent level (Caramazza, 1997; Costa,Caramazza, & Sebastian-Galles, 2000; Costa, Santeste-ban, & Cano, 2005; Dell, 1986; Dell, Schwartz, & Mar-tin, 1997; Griffin & Bock, 1998; Harley, 1993; Rapp &Goldrick, 2000; Starreveld & La Heij, 1995). According-ly, conceptual representations activated in the course oflexicalization activate their corresponding lexical nodes,which in turn spread some activation to their phonolog-ical content (see Fig. 1A).

In contrast, so-called discrete models restrict activa-tion flow in various different ways. There are two typesof discrete model. In Levelt, Roelofs and Meyer�s model(1999; see also Levelt, 2001), activation flows in acascaded fashion from the conceptual to the lexical level

(any activated conceptual representation spreads someactivation to the lexical level). However, phonologicalactivation is restricted to one lexical representation;the one that is selected for production (see Fig. 1B). Thatis, this model holds both the cascade assumption and thediscrete assumption: activation spreads in a cascadedmanner from the conceptual to the lexical system andin a discrete manner from lexical to sublexical systems.More recently, Bloem and La Heij (2003, see alsoBloem, van der Boogard & La Heij, 2004) proposed amodel in which only the conceptual representationincluded in the preverbal message (the one selected forproduction) passes activation to the lexical level (see alsoDamian & Bowers, 2003). However, the selected repre-sentation activates not only its lexical representationbut also those of semantically related items. In short,lexical activation, and as a consequence phonologicalactivation, is restricted (at maximum) to the target andsemantically related items (see Fig. 1C).1

These three proposals coincide in assuming that mul-tiple lexical representations are activated (at least thetarget along with semantically related items) in thecourse of lexical access. The agreement ends here. Thefull-cascade model and the discrete model proposed byLevelt et al. (1999) allow, in principle, for any activatedconceptual representation to send activation to its

2 Levelt et al. (1999) tried to accommodate Peterson andSavoy�s results by appealing to a malfunctioning of the lexicalselection mechanism. They argued that when two lexical itemsare very highly activated, as in the case of synonyms, the two ofthem may get wrongly selected and as a consequence the two ofthem activate their phonological codes. Thus, the phonologicalco-activation of synonyms is reflecting double lexical selectionrather than cascade processing. Despite the merits of suchexplanation, it is unclear whether it could also account for othereffects suggesting cascade processing and especially for thecognate effect observed by Costa et al. (2000). In this studybilingual speakers named pictures whose names varied onwhether their translations were phonologically similar (cog-nates) or dissimilar (non-cognates). Naming latencies werefaster for cognates than for non-cognates. The authors arguedthat this result reveals that there is phonological activation ofboth the target word in the response language and of itstranslation, supporting the notion of cascade processing. Anaccount of this effect in terms of double selection (e.g., selectionof the target word in the response language and also of itstranslation in the non-response language) seems highly unlikely.

E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377 361

corresponding lexical representations (regardless ofwhether they are semantically related to the intendedconcept). In contrast, in Bloem and La Heij�s model con-ceptual representations that are not included in the pre-verbal message do not activate their lexical nodes.Regarding which lexical representations are allowed tosend activation to the phonological level, full-cascade,and Bloem and La Heij�s models allow for the presenceof phonological activation of any activated lexical repre-sentation, while the discrete model proposed by Levelt etal. (1999) restricts phonological activation to the select-ed lexical node. Thus, the only model that keeps con-stant the governing principle throughout the system isthe full-cascade model. All other models are hybrids inthe sense that they assume that the flow of activationis guided by different governing principles dependingon the level of representation. Below we review someexperimental results that shed light on the flow of activa-tion in the speech production system.

Phonological activation of non-produced semantically

related words

Most of the studies addressing the flow of informa-tion in speech production have explored whether thereis phonological activation of conceptual representationsthat are semantically related to the target word.Although the first study addressing this question (Leveltet al., 1991; see also Jescheniak, Hahne, & Schriefers,2003) failed to observe phonological activation of non-selected lexical items, more recent studies have been ableto detect such phonological activation.

For example, Peterson and Savoy (1998) asked par-ticipants to perform a dual task experiment. Participantshad to name a set of pictures (e.g., couch), but on somecritical trials, they were asked to halt the naming processand instead name a target word printed on the screen.The printed word could be: (a) phonologically relatedto the picture�s name (e.g., count), (b) phonologicallyrelated to a near-synonym of the picture�s name (e.g.,soda which is related to �sofa�), (c) phonologically relatedto a semantically related word (e.g., bet which is relatedto �bed�), or (d) unrelated (e.g., harp). The results of thisexperiment replicated those of Levelt et al. (1991): (a)responses to words phonologically related to a semanti-cally related word (bet) were not different to unrelatedwords (harp), and (b) words phonologically related tothe picture�s name (count) led to faster latencies thanunrelated words. Importantly, participants namedwords phonologically related to the near-synonym�sname (soda) faster than unrelated words (harp) (see alsoJescheniak & Schriefers, 1998). This result was interpret-ed as revealing that during the retrieval of the target�sname, the phonological properties of both potential tar-get words (e.g., �couch� and �sofa�) are activated, leadingto the priming effect. The fact that phonological priming

was not observed for words phonologically related to asemantically related word (bet) was interpreted as sug-gesting that the phonological activation of a non-select-ed lexical node is only detectable when this node reachesa very high level of activation, as in the case of syn-onyms (see also Dell & O�Seaghdha, 1991, for a similarargument). This result supports the notion of cascadeprocessing and it inconsistent with the discrete modelproposed by Levelt et al. (1999).2

However, Peterson and Savoy�s (1998) results areconsistent with Bloem and La Heij�s discrete model,since this model allows for activation of semanticallyrelated representations both at the lexical and phonolog-ical levels. What this model forbids is activation of lexi-cal items (and hence activation of their phonology) thatare not semantically related to the target one. Thus, if wewant to adjudicate between Bloem and La Heij�s propos-al and the full-cascade models we need to evaluatewhether there is phonological activation of conceptualrepresentations that are not relevant for the communica-tive message and that are semantically unrelated to thetarget one.

Phonological activation of non-produced semantically

unrelated words

Several studies have explored whether there is activa-tion of conceptual representations that are not selectedfor production but that are at the same time semantical-ly unrelated to the target concept.

Cutting and Ferreira (1999) asked participants toname pictures while ignoring distractor words. In therelevant conditions, the pictures had homophone twins(e.g., a picture depicting a toy ball), and were presentedwith distractors that were semantically related to the

362 E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377

meaning of the target�s homophone twin (e.g., dance) orwith unrelated distractors (e.g., hammer). Naminglatencies were faster in the related condition than inthe unrelated condition. That is, distractors semanticallyrelated to the non-depicted meaning of the picture(dance) led to faster responses than unrelated distractors(hammer). This effect was explained as follows. The dis-tractor word dance activates its semantic representationalong with a cohort of semantically related concepts thatinclude BALL (dance). The activation of these conceptsspreads in a cascaded manner to the lexical system,thereby activating the lexical node �ball� (dance). Thisactivation further spreads to the phonological level,activating as a consequence the same phonologicalcontent of the target picture�s name [e.g., ball (toy)], thusspeeding up naming latencies. The authors� conclusionabout the results of this elegant experiment is perhapsthe clearest statement in favor of cascade processing:‘‘. . .phonological processing can be affected by semanti-cally processed material even when that material is notsemantically similar to the target representation’’ and‘‘Thus, unselected semantically processed material canaffect phonological processing, even when those lemmasare semantically dissimilar to the selected material(p. 334)’’ (see also Ferreira & Griffin, 2003, for conver-gent results).

Further support for this conclusion comes from astudy that used distractor pictures rather than distractorwords. Morsella and Miozzo (2002) presented partici-pants with two superimposed pictures and asked themto name one while ignoring the other. The two pictureswere presented in different colors, and the color servedas a cue signaling the target picture (e.g., name the pic-ture in green, ignore the picture in red). The distractorpicture was either phonologically related to the nameof the target picture (the target picture bell appearedalong with the distractor picture bed), or unrelated (thetarget picture bell appeared along with the distractorpicture hat). Naming latencies were faster in the relatedthan in the unrelated condition. The authors interpretedthis phonological facilitation effect (PFE) in a similarway as Cutting and Ferreira. They argued that wordsnon-selected for production, can nonetheless, andregardless of their semantic relationship with the target,activate their phonology.

These two studies provide strong evidence supportingthe full-cascade assumption, and are inconsistent withextant discrete models. However, there is also otherexperimental evidence that seems to be at odds withthe cascade model. We briefly discuss this evidence be-fore presenting our experiments.

First, in the context of a bilingual study, Costa,Miozzo, and Caramazza (1999) addressed the extent towhich a printed distractor word activates the phonolog-ical content of its translation in the response language.Participants were asked to name a picture in Catalan

[e.g., baldufa (spinning top)] while ignoring the presenta-tion of a distractor word in Spanish [e.g., pelea (fight)]whose translation in Catalan was phonologically similar[e.g., baralla (fight in Catalan)] to the target word. Alongthe same lines as Cutting and Ferreira�s rationale, theauthors argued that the distractor word (pelea) activatesits corresponding concept (semantically unrelated to thetarget baldufa), which in turn would activate its lexicalrepresentation in the response language (baralla). If cas-cade processing were to be functional, one may haveexpected the presence of phonological facilitation whenthe target and the distractor�s translation share somephonological features. However, the results did not sup-port this prediction, as naming latencies were unaffectedby this relationship.

Second, Bloem and La Heij (2003) failed to observephonological activation of distractor pictures in a simi-lar task to that used by Morsella and Miozzo (2002). Intheir experiment, participants were asked to translate aword from English (L2) into Dutch (L1) while ignoringthe presentation of a distractor picture. The authors ar-gued that if the distractor picture were to activate itsphonological content then translation times should befaster in the context of pictures whose names are pho-nologically related [e.g., borstel (brush)] to the targetword [e.g., bont (fur)] than in the context of unrelatedpictures [e.g., wiel (wheel)]. However, in conflict to Mor-sella and Miozzo�s observation, the predicted PFE wasabsent.

The third potential concern comes from the inconsis-tent results produced by semantically related distractorpictures (Bloem & La Heij, 2003; Damian & Bowers,2003; Glaser & Glaser, 1989; Humphreys, Lloyd-Jones,& Fias, 1995). The presence of phonological activationof a distractor picture implies the previous activationof its semantic and lexical representations. In such ascenario, one might expect to observe semantic effectswhen the distractor and target pictures are semanticallyrelated (see for example the semantic effects in thepicture–word interference paradigm, e.g., Lupker,1979; Rosinski, Golinkoff, & Kukish, 1975). However,the studies that have explored semantic effects of distrac-tor pictures lead to inconsistent results: semantic facilita-tion (Bloem & La Heij, 2003), semantic interference(Glaser & Glaser, 1989), and no effect at all (Damian& Bowers, 2003; Humphreys et al., 1995). Althoughdifferences in the experimental designs used in theseexperiments have been advanced as a possible cause ofthe discrepant results (Bloem & La Heij, 2003), at pres-ent the precise origin of the contrasting results is unclear(see General Discussion for a comparison between thesetasks).

The present experimental evidence does not give aconclusive answer to the question of whether semantical-ly unrelated conceptual information activates itscorresponding phonological content. In particular,

E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377 363

Bloem and La Heij�s (2003) results question the reliabil-ity of the PFE observed by Morsella and Miozzo�s(2002). Indeed, Bloem and La Heij (2003) argued thatthis PFE does not necessarily imply that any activatedconceptual representation spreads activation to the lexi-cal level. The reasoning of Bloem and La Heij is similarto that developed by Levelt et al. (1999) when account-ing for the presence of phonological activation of near-synonyms (Peterson & Savoy, 1998) (see Footnote 2).They argued that the PFE could be revealing a failurein the lexicalization process that selects for productionthe conceptual representation of the distractor picturerather than that of the target picture. As a result of thisfailure, the phonological properties of the distractorpicture will become activated. On these occasions, theyfurther argued, participants may have halted their lexi-calization processes before uttering the name of the dis-tractor picture and start the lexicalization of the targetpicture again. In this scenario, the retrieval of the pho-nological properties of the target word would be easierif part of these properties have been already pre-activat-ed by the distractor�s picture name (the phonologicallyrelated condition) than if they have not (the unrelatedcondition), leading to the presence of a PFE. Accordingto Bloem and La Heij the experimental conditionsused by Morsella and Miozzo are susceptible to suchderailments in the selection of the target representation,since target and distractors are difficult to discriminate.On this view, Morsella and Miozzo�s PFE does notreveal cascade processing but, rather a derailment inthe selection of the preverbal message.

The experiments reported in this article have severalaims. First, to assess the reliability of the PFE observedby Morsella and Miozzo (2002) using their same exper-imental setting. Second, to assess whether, under thesame experimental conditions in which PFE are ob-served, semantic effects can also be obtained. Third, toexplore whether the PFE for ignored stimuli is also pres-ent under experimental conditions that are not subject toBloem and La Heij�s criticism.

The rationale of our experiments is very similar tothat of Morsella and Miozzo (2002). If distractor pic-tures lead to a PFE then we can conclude that their cor-responding conceptual, lexical, and phonologicalrepresentations have been activated during the courseof lexical access.

Experiment 1: Contextual effects in picture naming from

distractor pictures

The main goal of this experiment is to evaluate the ef-fect of phonologically related and semantically relateddistractor pictures under the same experimentalconditions in a picture naming task. In this experiment,participants were presented with two superimposed

pictures (one colored in green and the other colored inred) and were asked to name the green picture whileignoring the red one.

Experiment 1a: Phonological effects from distractor

pictures

Method

Participants. Thirty-six native speakers of Spanish, stu-dents at the University of Barcelona, took part in theexperiment in exchange for a course credit.

Materials. Twenty-four pictures were used as targetpictures and another set of 24 pictures was used as dis-tractors (line-drawings taken mostly from the Snod-grass & Vanderwart�s set, 1980). Each target picture[e.g., boca (mouth)] appeared along with a distractorpicture whose name was phonologically related [e.g.,bota (boot)], and along with a distractor picture whosename was phonologically unrelated [e.g., lapiz (pencil)].Furthermore, and in order to reduce the number ofrelated items, target pictures also appeared withanother set of 24 filler distractor pictures that wereunrelated. Thus, the target pictures appeared threetimes each: once with a related distractor and twicewith unrelated distractors. The names of the picturesincluded in the phonologically related condition sharedan average of 2.3 segments and always shared at leastthe first two segments (see Appendix A). Target pic-tures appeared in green and distractor pictures inred. The pictures of each pair appeared simultaneouslyand were superimposed.

To further reduce the number of related trials, a sec-ond set of 24 filler target pictures was presented threetimes along with a distractor picture. None of these fillertarget and distractor pictures was used in the experimen-tal conditions. In total, there were 48 target pictures (24experimental + 24 filler) that appeared three times each;and 72 distractor pictures (24 experimental + 48 fillers)that appeared two times each.

In the overall experiment, each participant was pre-sented with 48 experimental trials (24 trials in the relatedcondition and 24 trials in the unrelated condition) and96 filler trials, all of them unrelated. In this way, the per-centage of related trials is rather low (16%). The exper-iment contained three different blocks of 48 trials each.Target and distractor pictures appeared only once perblock, and the two experimental conditions were distrib-uted equally across the blocks (eight times per block).Trials inside each block were randomized with therestriction that two phonologically related trials ap-peared with a minimum distance of three trials betweenthem. The first two trials at the beginning of each blockcontained filler pictures. Care was taken to avoid anyobvious relationship (semantic or phonological) betweenthe pictures of two successive trials in order to prevent

Table 1Average naming latencies (Mean), standard deviations (SD)and error rates (E%) broken by condition for Experiment 1a

Type of relationship Language

Spanish English

Mean SD E% Mean SD E%

Phonologically related 737 73 4.7 744 80 4.4Unrelated 758 73 4.9 749 85 4.3

Phonological effect � 21 �5

364 E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377

the emergence of negative priming (e.g., Damian, 2000;Tipper, 1985). Participants were randomly and equallyassigned to six different block orders.

Procedure. Participants were tested individually in asound-attenuated room seated approximately 60 cmfrom the screen. At the beginning of the experiment,participants were presented with the 48 target pictures(without distractors) and were instructed to name them.Afterwards, a training phase started in which the targetpictures appeared along with unrelated distractor pic-tures. None of these unrelated pictures were includedin the experimental session. Participants were asked toname the pictures that appeared in green (the targetones) as fast and accurately as possible, while ignoringthe pictures that appeared in red (the distractor ones).An experimental trial involved the following events: (a)a fixation point (an asterisk) was shown in the centerof the screen for 1250 ms; (b) a blank interval of500 ms; (c) the picture–picture stimulus was presenteduntil subject�s response or for 800 ms; (d) 2000 ms afterthe response or after the onset of the stimulus the trialterminated; (e) a question mark appeared and a new trialbegan after participants pressed the spacebar. Responselatencies were measured from the onset of the picture–picture presentation. To check that distractor pictureselicited the expected name each participant was askedto name them after the experimental session. Given thatthese pictures were selected on the basis of their highname agreement, it is not surprising that all subjects pro-duced the expected name for all pictures. Stimulus pre-sentation and reaction times were controlled by theDMDX program (Forster & Forster, 2003). The entireexperimental session lasted for approximately 35 min.

Results

Three types of responses were excluded from theanalyses: (a) production of names that differed fromthose designated by the experimenter; (b) verbal disflu-encies (stuttering, utterance repairs, and production ofnonverbal sounds that triggered the voice key); and (c)recording failures. Also, naming latencies below300 ms or above 3 SD from a given participant�s meanwere discarded from the analyses (4.8% of the datapoints were excluded). Error rates and naming latenciesin the phonologically related condition were comparedto those in the unrelated condition (see Table 1).

No significant differences were observed in the anal-ysis of error rates (ts < 1). However, naming latencies inthe phonologically related condition were 21 ms fasterthan in the unrelated condition (t1(35) = 5.15; p < .01;t2(23) = 2.04; p < .06), replicating the PFE from dis-tractor pictures reported by Morsella and Miozzo(2002).

The result of this experiment suggests that a phono-logical overlap between the name of the ignored picture

and the name of the target picture facilitates naminglatencies. However, before reaching such a conclusionit is important to assess whether the factor behind thePFE is actually the phonological overlap between thetarget and distractor, and no other uncontrolled vari-ables (e.g., visual masking). To this end we asked agroup of 36 native speakers of English, students at Har-vard University, to participate in the same experiment inEnglish. Crucially, here, phonological overlap betweendistractor and target names was absent in the two condi-tions. The results of this control group showed no signif-icant differences across conditions (all ts < 1), revealingthat the difference observed in Experiment 1a is actuallydue to the phonological overlap between the names ofthe distractors and the target names (see Table 1).

Experiment 1b: Semantic effects from distractor pictures

As reviewed in the Introduction, the current experi-mental evidence on the effects of semantically relateddistractor pictures in picture naming is mixed. Semanti-cally related distractors has led to semantic interference(Glaser & Glaser, 1989) and to null effects (Damian &Bowers, 2003). However, neither of these studies testedsemantic and phonological effects under the same exper-imental conditions. Thus, we cannot safely conclude thatthere are no effects of semantically related distractorsunder the same experimental conditions in which phono-logical facilitation is observed. Experiment 1b aims atresolving this uncertainty.

The details of this experiment are very similar tothose of Experiment 1a. However, given the elusive nat-ure of semantic effects produced by distractor pictures,we wanted to make sure that our experiment was sensi-tive enough to detect an effect. Thus, we included a con-dition in which the target picture appeared without adistractor. Presumably, naming latencies should be fast-er when the picture appears in isolation than when itappears along with a distractor.

Method

Participants. Eighteen participants from the same pop-ulation as in Experiment 1a took part in this experiment.

E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377 365

Materials. The design of this experiment was very sim-ilar to the previous one with the following modifica-tions. First, the 48 target pictures (24 target, 24 filler)appeared twice along with distractor pictures and oncein isolation, for a total of 144 trials. Also, target anddistractor pictures in the related condition belong tothe same semantic category. When comparing the ef-fects of semantically related distractor pictures againstunrelated distractor pictures it is important to controlthe visual similarity between the target and the distrac-tors in the two conditions. Several measures were takento ensure that the visual similarity between related andunrelated picture pairs was similar. First, we avoidedpairing two objects with very obvious visual overlap(e.g., table and stool were not paired). Second, we con-ducted a norming study in which visual similarity rat-ings were gathered. In this study, each target picturewas paired with several distractor pictures (relatedand unrelated). The resulting 130 object pairs were pre-sented to 23 participants, who were asked to rate thevisual similarity between the two objects of the pairpresented side by side (1: not similar at all; 5: very sim-ilar). Based on these ratings we selected objects to bepaired with each target picture. When carrying outthe selection process, we tried to equate the visual sim-ilarity between target and distractor as much as possi-ble in the two conditions (visual similarity for relatedobjects: 1.73; visual similarity for unrelated objects:1.65; t < 1) (see Appendix B).

Results

Following the same criteria as in Experiment 1a, 5.8% of the data points were excluded from the analyses.

Semantically related distractor pictures elicited asmany errors as unrelated ones (all ts < 1). Also, naminglatencies were not affected by the semantic relationshipbetween target and distractor (all ts < 1). Finally, nam-ing latencies were slower when the target picture ap-peared along with a distractor picture than in isolation(all ps < .05) (see Table 2).

The results of this experiment fail to show any mea-surable effect of semantically related distractor picturesin picture naming. However, before concluding thatsemantic effects are absent in this task it is important

Table 2Average naming latencies (Mean), standard deviations (SD),and error rates (E%) broken by condition for Experiment 1b

Type of relationship Picture–picture Picture–word

Mean SD E% Mean SD E%

Semantically related 763 77 5.8 777 71 5.8Unrelated 762 73 6.3 754 58 3.2Isolated picture 665 67 5.3 646 56 2.3

Semantic effect 1 23

to assess whether the distractors were able to elicitsemantic effects at all. To this end we carried out a pic-ture–word interference experiment, in which the namesof the distractors were presented visually and partici-pants were instructed to name the target pictures. In thisparadigm, categorically related distractors usually leadto semantic interference (e.g., Lupker, 1979; Rosinskiet al., 1975). The results of this latter experiment(n = 18) revealed a reliable 23 ms. semantic interference(t1(17) = 2.76; p < .02; t2(23) = 1.89; p < .08). Thus, thesemantic relationship held by the target and distractorswas strong enough to lead to measurable semantic ef-fects in an experiment with the same number of partici-pants (see Table 2).

Discussion Experiment 1

Two main observations were made in Experiment 1:(a) a phonological relationship between the name of adistractor picture and the name of a target picturespeeds up naming latencies (Experiment 1a), (b) asemantic relationship between the two stimuli does notaffect naming latencies (Experiment 1b).

The PFE replicates Morsella and Miozzo�s observa-tion, and strongly suggests that during picture namingdistractor pictures activate their phonological code,hence supporting full-cascade models of lexical access.The failure to obtain semantic contextual effects from ig-nored pictures also replicates recent observations(Damian & Bowers, 2003; Humphreys et al., 1995). Inthe Introduction we argued that the absence of semanticeffects in the context of phonological effects might seemsurprising. Our contribution here is that under the sameexperimental conditions such a pattern is reliably ob-tained. In the following we focus on the PFE, and we de-fer further discussion of the absence of semantic effectsto the General Discussion.

As we argued, the PFE is inconsistent with the twodiscrete models presented in the Introduction. Accord-ing to Levelt et al.�s (1999) model only one lexical itemis phonologically encoded; the one that is selected forproduction. Given that the distractor picture is neverselected or produced, its phonological content shouldnot be activated. According to Bloem and La Heij�smodel, phonological activation is restricted to the targetlexical node and to semantically related items. There-fore, no phonological activation of semantically unrelat-ed pictures should be present.

However, as advanced in the Introduction, the pres-ence of phonological co-activation of distractor pic-tures in picture naming might stem from an error inthe decision of which picture to lexicalize. The nextexperiments aim at exploring the presence of phonolog-ical activation of non-intended conceptual representa-tion while reducing the chances of a derailment inthe lexicalization process. We do so by: (a) making

366 E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377

the target and distractor dimensions easier to discrimi-nate at the physical level, and (b) limiting the responseset to one type of conceptual representation (color con-cept) which is different from the ignored one (objectconcept).

Experiment 2: Phonological activation of distractor

pictures in color naming

In Experiment 2 participants were presented with col-ored pictures and they were asked to produce the namesof the colors in Spanish (e.g., the picture candle appearsin brown, and participants have to say ‘‘brown’’). In thissituation, participants need to retrieve neither the con-cept of the target picture, nor its lexical representationor phonological content. Furthermore, the conceptualdimension that needed to be lexicalized is clearly differ-ent from that needed to be ignored, making the chancesof an incorrect conceptual selection highly improbable.This is because participants know in advance that theywill be naming only colors. In some cases the name ofthe object (distractor dimension) was phonologicallyrelated to the name of the color (target dimension).For example, in the phonologically related conditionthe object vela (candle in Spanish) appeared in verde

(green in Spanish), while in the unrelated condition ap-peared in marron (brown in Spanish). If the phonologi-cal content of the depicted object (distractor dimension)were activated, then naming latencies in the phonologi-cally related condition would be faster than in the unre-lated condition. Alternatively, if the phonologicalactivation is restricted to the selected conceptual (or lex-ical) representation (i.e., the color) then color naminglatencies should be independent of the phonologicalproperties of the object�s name.

Two groups of participants were included in thisexperiment. Participants in Group 1 were asked to namethe color of the picture by means of a gender-markedutterance in Spanish: gender-marked determiner + nameof the color, e.g., ‘‘la verde’’ [literally ‘‘the (fem) green’’].In Spanish, some closed-class words such as pronounsand determiners depend on the gender of the noun. Forexample, when referring to a picture of a candle as ‘‘thegreen one’’ the corresponding utterance in Spanish car-ries a gender-marked determiner, ‘‘la verde’’ [literally’’the (fem) green’’]. The determiner referring to femininenouns is la and the one referring to masculine nouns is el.This group was included to ensure that the paradigm andmaterials were sensitive to the presence of phonologicaleffects. It is generally assumed that in order to retrievethe correct determiner form (la for feminine nouns, andel for masculine nouns), participants need to retrievethe lexical representation of the noun along with its gen-der value. Thus, given that the selection of the noun�s lex-ical representation is needed to produce utterances such

as ‘‘la verde,’’ we should expect the phonological contentof such a noun to be activated (Schmitt, Meyer, & Levelt,1999, see however Jescheniak, Schriefers, & Hantsch,2001). As a consequence, we should observe phonologi-cal effects for this group of participants.

More interesting for our purposes is the performanceof participants in Group 2, who were asked to respondusing only the name of the color in which the picturewas depicted (e.g., ‘‘verde,’’ green). The predictions ofthe different models for this group of participants paral-lel those of Experiment 1a.

Method

Participants

Forty-four participants from the same population asin Experiment 1 took part in this experiment.

Materials

The selection of the materials was constrained by thereduced number of picturable objects that have a phono-logical overlap with color names in Spanish. Also, wewanted to avoid the use of objects with obvious naturalcolors. We selected four pairs of objects [e.g., vela (can-dle)-ventana (window); nariz (nose)-navaja (clap knife);roca (rock)-rodilla (knee); maleta (suitcase)-mariposa

(butterfly)]. The names of the objects in each pair hada phonological overlap with one of the four colorsincluded in the experiment [verde, naranja, rojo, andmarron (green, orange, red, and brown respectively)].For example, the object mariposa (butterfly) and maleta

(suitcase) were phonologically related to marron

(brown), while the objects vela (candle) and ventana

(window) were related to verde (green).The eight objects appeared in each of the four colors

included in the experiment, leading to 32 different targetpictures. Only for eight color-object combinations werethe names of the color and the object phonologicallyrelated. In the related condition, object and color namesshared an average of 2.1 segments, and shared at leasttheir first two segments (see Appendix C).

All experimental objects were of feminine gender. Tobalance the number of times participants had to producefeminine and masculine gender-marked utterances (forGroup 1) we selected another set of eight pictures withmasculine names. These pictures appeared along withthe four colors used in the experiment. No phonologicaloverlap between these pictures and the colors was pres-ent and we thus considered these stimuli as fillers. Also,the inclusion of these filler trials reduced the percentageof related trials to 12.5%. Overall participants were pre-sented with 64 stimuli (16 objects that appeared each inthe four colors included in the experiment). However, inorder to gain more experimental power, the 64 itemswere presented twice leading to a total of 128 trials. Par-ticipants were presented with two blocks of 64 items

Table 3Average naming latencies (Mean), error rates (E%), andstandard deviations (SD) broken by type of utterance, exper-imental condition, and language for Experiment 2

Type of relationship Spanish Catalan

Mean SD E% Mean SD E%

Determiner + color naming

Phonologically related 784 69 9.9 826 137 10.4Unrelated 800 71 11.3 828 128 10.8

Phonological effect �16 �2

Color naming

Phonologically related 540 61 1.1 551 60 3.9Unrelated 561 57 4.1 556 65 4.4

Phonological effect �21 �5

E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377 367

each (eight different blocks were constructed). All 64 ob-ject-color combinations were therefore present in eachblock. The order of the stimuli presentation in eachblock was randomized with the following restrictions:(a) stimuli from the related condition were separatedby at least four trials, (b) stimuli containing the same ob-ject were separated by at least three trials, and (c) succes-sive trials containing the same color were avoided. Thefirst two stimuli of each block were always filler stimuli.Each participant received two different blocks. A givencombination of two blocks was never assigned to morethan one participant.

Procedure

At the beginning of the experiment, participants werepresented with the entire set of objects in black andwhite along with their written names and were instruct-ed to name them aloud with the proper determiner form(e.g., ‘‘la vela,’’ the candle). Afterwards, participantswere informed that they would see the same objectsbut in various colors (green, red, brown, and orange).Participants in Group 1 were instructed to respond toeach picture by means of a determiner + color utterance[e.g., ‘‘la verde,’’ the (fem) green; ‘‘el rojo,’’ the (mas)red]. Participants of Group 2 were instructed to namethe color in which the object was depicted while ignor-ing the meaning of the object (e.g., ‘‘verde,’’ green;‘‘rojo,’’ red). After a training block containing the 64items, the experiment proper started. Each trial hadthe following structure: (a) a fixation point (an asterisk)was shown in the center of the screen for 1000 ms, fol-lowed by a blank interval of 450 ms, (b) the colored pic-ture was presented on the center of the screen untilsubject response or for 800 ms, (c) a question mark ap-peared on the screen 1500 ms after the picture�s disap-pearance; (d) the next trial began after the participantpressed the spacebar. Response latencies were measuredfrom the onset of the stimulus to the beginning of thenaming response. The experiment was controlled byEXPE software (Pallier, Dupoux, & Jeannin, 1997). Re-sponse latencies were measured by means of a voice key.The session lasted for about 35 min.

Results

Following the same criteria as in Experiment 1a,7.1% of the data points were excluded from the analyses.Two variables were analyzed: ‘‘Phonological Relation-ship’’ (Related vs. Unrelated), ‘‘Type of Utterance’’(Determiner + Color vs. Color). Given the limited num-ber of items (four colors) we did not carry out an itemanalyses.

In the error analyses the only significant effect wasthat of the variable ‘‘Type of Utterance’’ (F (1,42) =71.22; MSE = 1408.66; p < .01), revealing that partici-pants made more errors in the ‘‘Determiner + Color’’

utterance than in the ‘‘Color’’ utterance. The effectof ‘‘Phonological Relationship’’ was non-significant(F < 1) (see Table 3).

In the naming latencies analyses, the main effect of‘‘Phonological Relationship’’ was significant (F (1,42)= 23.26; MSE = 7415.69; p < .01), revealing that nam-ing latencies were faster in the phonologically relatedcondition than in the unrelated condition. The maineffect of the variable ‘‘Type of Utterance’’ was also sig-nificant (F (1,42) = 158.48; MSE = 1279820; p < .01),revealing slower latencies for the ‘‘Determiner + Color’’utterance than for the utterance ‘‘Color.’’ More impor-tantly, the interaction between these two conditionswas not significant (F < 1), revealing that the differencebetween the related and unrelated conditions was statis-tically similar in both utterance formats (16 and 21 msfor ‘‘Determiner + Color’’ and ‘‘Color,’’ respectively).

In this experiment, naming latencies were faster whenthe name of the depicted object was phonologicallyrelated to the name of the color, suggesting that the pho-nological properties of the depicted object were activat-ed in the course of color naming. However, as we arguedin Experiment 1, before attributing such an effect to thephonological activation of the ignored stimuli, we needto be sure that it is not due to other uncontrolled vari-ables (perhaps, the object-color combinations in thephonologically related condition were more familiarthan in the unrelated condition). Following the samerationale as in Experiment 1a, we addressed this concernby comparing the same object-color combinations in anexperimental situation in which the phonological over-lap is absent. We asked native speakers of Catalan toperform the same task with the same materials as inExperiment 2, but in Catalan. Crucially, in this languagethe names of the colors and objects were phonologicallyunrelated. Forty-four native speakers of Catalan fromthe same population as in Experiment 1 took part in thisexperiment. Half of the participants were assigned to

368 E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377

Group 1 (Determiner + Color utterance format) and theother half to Group 2 (Color utterance format). Weexcluded the data points for the item maleta (suitcase)because its name in Catalan was phonologically relatedto the color name marro (brown). The results of this con-trol experiment did not show any significant differencebetween the phonologically related and unrelated condi-tions (F < 1) (see Table 3).

Discussion

The facilitation effect reported in the ‘‘Color’’ condi-tion in Experiment 2, in which participants had to pro-duce only the name of the color in which an objectwas depicted, suggests the existence of phonological acti-vation of a stimulus that is irrelevant for the lexicaliza-tion process (the name of the depicted object).Crucially, when the task was conducted in Catalan, nosuch result emerged. This indicates that the differencebetween two conditions observed in Experiment 2 wasdue to the phonological overlap between object and col-or names in the related condition. This pattern of resultsis consistent with the observations made in Experiment1a (see also Morsella & Miozzo, 2002) and supportsthe notion that the flow of activation in speech produc-tion honors the cascade principle. In the next two exper-iments we further test the reliability of the PFE in otherexperimental contexts.

In Experiment 3, we assess the impact that thefamiliarization phase and the extensive repetition ofthe stimuli may have had on the presence of the PFEobserved in Experiment 2. In the latter, participantswere familiarized with the names of the ignored pic-tures before the color naming task. One could arguethat such familiarization could induce the retrieval ofthe to-be-ignored object name during the experimentalphase (i.e., color naming), hence leading to the ob-served PFE. Also, in Experiment 2, the to-be-ignoredpictures were repeated many times during the experi-mental session (12 times). It is possible that this exten-sive repetition of the to-be-ignored items enhances thechances for detecting a PFE,3 an effect that under morenatural conditions might be absent. Experiment 3addresses the impact of these two variables in thedetectability of the PFE.

3 Note that if it were to be the case that the PFE arises as aconsequence of the extensive repetition of the pictures, the onlymodel that would be able to account for the PFE would still bethe cascaded model. This is so because discrete models wouldnot predict activation of the object names even after manyrepetitions and, in this way, these models could not account forthe PFE reported in Experiment 2.

Experiment 3: The impact of familiarization and

repetitions on the presence of the PFE

This experiment is very similar to Experiment 2 butwith two major modifications: (a) participants werenot familiarized with the names of the to-be-ignored pic-tures before the experimental session and (b) each pic-ture was presented only five times (in comparison to12 times in Experiment 2). If the PFE observed in Exper-iment 2 stems from the cascaded nature of the speechproduction system we should observe it also in the pres-ent experiment.

Method

Twenty-two participants from the same populationas in Experiment 2 took part in the experiment. All ofthem were instructed to name the color in which the ob-ject was depicted. The same materials as in Experiment 2were used here. Unlike in Experiment 2, participantswere not familiarized with the names of the to-be-ig-nored pictures. The training phase included only 16 tri-als, in which each object appeared only once, and eachcolor four times. No related object-color combinationswere presented in this phase. After the training phasethe main experiment began. Each participant was pre-sented with 64 trials. Stimuli presentation and blockswere the same as in Experiment 2 (however each partic-ipant was only presented with one block rather thanwith two as in Experiment 2). To check that the distrac-tor objects elicited the expected name, each participantwas asked after the experimental session to name theobjects.

Results and discussion

Following the same criteria as in Experiment 2,4.5% of the data points were excluded from the anal-yses. Before submitting the data to the statistical anal-yses, we checked, for each participant (by assessingtheir performance in the naming task conducted afterthe experiment), whether the to-be-ignored pictureelicited the expected name. For those items in whichthis was not so, we removed the corresponding nam-ing latencies (13.5%). In total, 18% of trials werediscarded from analyses. The same data analyses asin Experiment 2 were conducted here, but includingonly one independent variable, ‘‘PhonologicalRelationship.’’

In the error analyses no differences between the twoconditions were observed (t < 1). In the naming latenciesanalyses the effect of the variable ‘‘PhonologicalRelationship’’ was significant (t (21) = 2.38; p < .03),revealing that naming latencies in the related conditionwere 24 ms faster than in the unrelated condition (seeTable 4).

Table 5Average naming latencies (Mean), error rates (E%), andstandard deviations (SD) broken by experimental conditionfor Experiment 5

Type of relationship Color patch naming

Mean SD E%

Phonologically related 548 60 4Phonologically unrelated 560 60 5.2

Phonological effect �12

Table 4Average naming latencies (Mean), error rates (E%), andstandard deviations (SD) broken by experimental conditionfor Experiment 4

Type of relationship Color naming

Mean SD E%

Phonologically related 581 130 4.5Phonologically unrelated 604 147 4.5

Phonological effect �24

E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377 369

The results of this experiment replicated the PFEobserved in Experiment 2. That is, naming latencieswere faster when the name of the to-be-ignored picturewas phonologically related to the name of the colorthan when it was not. The fact that the two experi-ments differed in the: (a) presence of familiarizationphase and (b) extensive repetition of the to-be-ignoredpictures, but that nevertheless the PFE is observed inboth, suggests that neither of these factors is crucialfor the detectability of the effect.4 Table 5. To recapit-ulate, the PFE reported in Experiments 1a, 2, and 3strongly suggests the existence of phonological activa-tion of a stimulus that is irrelevant for the lexicaliza-tion process (the name of the depicted object) in thecourse of lexical access. These results support thenotion that the flow of activation in speech productionhonors the cascade principle. Experiment 4 further teststhis hypothesis in a slightly different experimentalcondition that minimizes the chances that participantsmisselect the target dimension.

4 Further support for this conclusion comes from areanalysis of Experiment 2, in which we assessed themagnitude of the PFE across the two blocks included inthe experiment. The magnitude of the PFE was identical inboth blocks (First block: phonologically related condition:542 and unrelated condition: 563; second block: phonologi-cally related condition: 539 and unrelated condition: 560).Furthemore, these magnitudes were similar to that observedin Experiment 4 (24 ms).

Experiment 4: A further test of phonological activation of

distractor pictures

In this experiment, we make the selection of the tar-get representation easier, by physically uncoupling thetarget and distractor dimensions: participants have toname a color patch that appears in the middle of thedepicted object. By physically uncoupling the attendedand the irrelevant dimension we minimize the chancesthat participants misselect for production the irrelevantdimension (the depicted object). The same objects as inExperiment 2 were presented in this experiment, butdepicted in black and white, and with a superimposedcolor patch. Participants were instructed to name thecolor patch and ignore the depicted object.

Method

Twenty-two participants from the same populationas in Experiment 1 took part in the experiment. Objectswere presented along with a color opaque rectangle of2 · 0.8 cm. The rectangle appeared superimposed onthe middle of the picture. For one given picture the rect-angles of the four different colors appeared always in thesame position. All other details are identical to those inExperiment 2.

Results and discussion

Following the same criteria as in Experiment 2, 4.9%of the data points were discarded from the analyses. Thesame data analyses as in Experiment 2 were conductedhere, but declaring only one independent variable, ‘‘Pho-nological Relationship.’’

In the error analyses no differences between the twoconditions were observed (t < 1). Naming latencies werefaster (12 ms) in the related than in the unrelated condi-tion (t (21) = 2.02; p < .06).

Given the similarities between Experiments 2 and 4we conducted a joint analysis in which we declared awithin-subjects variables, ‘‘Phonological Relationship’’(Related vs. Unrelated), and a between subjects variable,‘‘Type of Utterance’’ (Determiner + Color, Color, ColorPatch).

In the error analysis, themain effect of ‘‘Type ofUtter-ance’’ was significant (F (1,63) = 41.85; MSE = 763.72;p < .01). Participants made more errors in the unrelatedthan in the related condition (F (1,63) = 3.7;MSE = 110.61; p < .06). The interaction between thetwo variables was not significant (F < 1).

In the naming latencies analysis, the main effect ofthe variable ‘‘Type of utterance’’ was significant(F (1,63) = 109.68; MSE = 843729.91; p < .01). Themain effect of the variable ‘‘Phonological Relationship’’was also significant (F (1,63) = 25.95; MSE = 8414.34;p < .01), revealing faster naming latencies in the related

370 E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377

condition than in the unrelated condition. Importantly,the interaction between these two variables was notsignificant (F < 1), revealing that the phonologicalfacilitation effect is comparable in the three utteranceformats.

The results of Experiment 4 replicated the PFEproduced by ignored distractor objects observed inExperiments 1a, 2, and 3, suggesting that distractorobjects activate their corresponding phonological formin the course of lexicalization.5 Importantly, this effectis present even under experimental conditions in whichthe target dimension and the distractor dimension arephysically uncoupled.

General discussion

We reported four experiments assessing the effects ofdistractor objects during naming. The main objective ofthis series of experiments was to explore the extent towhich distractor objects activate the phonological con-tent of their names in the course of lexicalization.

In Experiment 1a, participants named pictures(depicted in green) while ignoring the presentation ofsuperimposed distractor pictures (depicted in red). Nam-ing latencies were faster when the distractor�s picturename was phonologically related to the name of thetarget picture than when it was unrelated, replicatingprevious observations by Morsella and Miozzo (2002).In Experiment 1b, a semantic relationship between thetwoobjects did not affect naming latencies (see alsoDami-an & Bowers, 2003). In Experiment 2 participants wereinstructed to name the color in which an object wasdepicted. Naming latencies were faster when the nameof the target color was phonologically related to the ob-ject�s name. However, such an effect was not present whenthe task was performed in a language in which no phono-logical relationship between the paired color-object

5 Convergent evidence supporting this explanation comesfrom a reanalysis of the results of the color-naming experi-ments. If the PFE originates from the phonological activationof the distractor picture name, then one might expect a largerPFE for those trials in which color-naming latencies are slow.This is because, if color-naming latencies are very fast,phonological activation of the distractor picture name mayarrive too late to affect the retrieval of the phonologicalproperties of the target word, and hence the PFE would beabsent. To test this prediction, we conducted a reanalysis inwhich the RTs (for each subject and condition) were split intotwo halves (faster and slower color-naming RTs). The resultsconfirmed this prediction and the PFE was only reliablyobserved for the slower half (Magnitude of the PFE: FastNaming latencies: �9, �1, and �1 ms/Slow Naming Latencies:�33, �47, and �21 ms; for Experiments 2, 3, and 4, respec-tively). We thank Dr. La Heij for drawing our attention to thisaspect of the results.

names was present (Catalan), suggesting that the PFE ob-served in Experiment 2 was actually due to the phonolog-ical relationship between the color and the objects� names.Experiment 3 revealed that neither a familiarizationphase nor a extensive repetition of the experimental pic-tures is responsible for the presence of the PFE in Exper-iment 2. Finally, in Experiment 4 the PFE was observedunder experimental conditions inwhich the discriminabil-ity of targets and distractors was enhanced.

The presence of the PFE produced by irrelevant pic-torial stimuli in various experimental contexts highlightsthe reliability and reproducibility of the phenomenon.We were able to observe such an effect under four slight-ly different experimental conditions. Furthermore, itmakes an explanation of the PFE in terms of an errorin the selection of the proper conceptual representationfor lexicalization highly unlikely. Instead, the PFEstrongly suggests that the phonological properties of pic-torial stimuli that do not need to be lexicalized (theyactually need to be ignored) become activated in thecourse of naming. This observation has important impli-cations for models of lexical access in speech productionand in particular for the processing dynamics acrosslevels of representation.

In the Introduction we discussed three different pro-posals regarding the flow of activation across the differentlevels of representation in speech production (the concep-tual, the lexical, and the phonological levels). The maindifference between them is the extent to which they allowactivation to spread freely across these levels. Bloem andLaHeij�s (2003) proposal assumes that only the conceptu-al representation included in the preverbal message pass-es activation to the lexical system. This conceptualrepresentation activates its corresponding lexical repre-sentation along with a cohort of semantically related lex-ical items. As a consequence, phonological activation isrestricted, at maximum, to the target lexical item andsemantically related ones. The other model that restrictsthe flow of activation across levels of processing is thatproposed by Levelt et al. (1999), where only the selectedlexical representation activates its phonological form.Despite the differences between these two proposals, theyboth predict that conceptual information that is not partof the preverbal message (and that is not semanticallyrelated to it) should not activate its phonological content.Our results are at odds with this prediction.

However, the presence of phonological activation ofdistractor pictures in the course of lexicalization ispredicted by models that assume a free spread ofactivation across different processing levels (Caramazza,1997; Dell, 1986; Dell et al., 1997). According to thesecascade models, any activated representation spreadsproportional activation to other representations withwhich they are linked. Thus, if a conceptual representa-tion during speech production is activated (e.g., via thepresentation of a distractor picture), then such a

E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377 371

representation would spread some of its activation tosubsequent levels of processing, reaching, to someextent, the phonological level. This would be so evenfor conceptual representations that are not relevant tothe lexicalization process (e.g., they are not included inthe preverbal message) and are unrelated to the targetone. Therefore, the results reported in our experimentssupport the notion that activation flows in a cascademanner through the whole speech production system.6

The contribution of our study is the demonstration ofreliable phonological effects from ignored pictures invarious experimental naming contexts. However, webelieve that for the sake of completeness, it is necessarythat we attempt to reconcile the presence of this phono-logical activation with some experimental observationsthat might seem, at first sight, inconsistent. The firstrefers to the presence of phonological effects in a namingexperimental context in which semantic effects are notpresent. The second refers to the contrasting resultsobserved with relatively similar paradigms (e.g., transla-tion tasks). These issues are discussed below.

The presence of phonological effects in the context of no

semantic effects

In the Introduction we advanced a seemingly para-doxical observation: the presence of phonological effectsin the same context in which semantic effects are absent.Indeed, the results of our Experiments 1a and 1b con-tribute to further reaffirm the reliability of such a patternof results (Damian & Bowers, 2003; Humphreys et al.,1995). At first sight, one may be tempted to predict thatin those experimental circumstances in which there isphonological activation of a distractor picture, somesort of semantic effects should also be observed whenthe target and the distractor hold a semantic relation-ship. This is because, for the phonological propertiesof the distractor to become activated, their correspond-ing conceptual and lexical representations need to havebeen activated previously. But does such a predictionnecessarily follow from the presence of phonologicalactivation of distractor pictures? We think not.

As stated above, when accounting for the PFE one isforced to assume that the conceptual and lexical repre-sentations of the distractor picture are activated. Given

6 We have discussed the implications of the PFE in thecontext of feed-forward not-interactive models of lexical access.However, there are several proposals in the literature arguingthat the speech production system entails some interactiveprocessing (Dell, 1986; Harley, 1993; Rapp & Goldrick, 2000),in the sense that activation of phonological representationsfeeds-back to higher lexical representations. The presence of thePFE is completely consistent with interactive models. In fact,the PFE could be revealing the contribution of these twoprinciples. And, in fact, all interactive models embrace to someextent the cascade principle.

the activation of these two types of representation, it isthen appropriate that we consider the effects that asemantic relationship may have at both of these levelsof processing.

A semantic relationship between target (e.g., lion)and distractor picture (e.g., tiger) may help the retrievalof the conceptual representation of the target picture(Damian & Bowers, 2003, see below). That is, recogni-tion of the target picture (or selection of its conceptualrepresentation) would be faster in the context of asemantically related picture than in the context of anunrelated one, because of the priming exerted by therelated distractor (see Bloem & La Heij, 2003; for thesame argument). Why then, is no semantic facilitationobserved for distractor pictures in the majority ofpicture naming experiments?

If we assume cascade processing, the semantic repre-sentation of both the target (e.g., lion) and the distractor(e.g., tiger) would spread some activation to their corre-sponding lexical representations. At this level of process-ing there is wide agreement in assuming that the easewith which a lexical representation is selected dependson its level of activation in relation to that of other acti-vated lexical representations that act as competitors(e.g., Caramazza & Costa, 2000; Levelt et al., 1999;Roelofs, 1992). The larger the discrepancy in the levelof activation of the target and that of the competitors,the easier lexical selection is. Thus, the selection of thetarget lexical node �lion� would depend not only on itslevel of activation but also on the level of activation of�tiger� in the related condition and of �chair� in the unre-lated condition. Presumably, the activation-level of arelated distractor �tiger� would be larger than that ofan unrelated one �chair� because of the conceptualoverlap between the conceptual representations of theformer distractor and the target (LION). In this scenar-io, lexical selection would be harder in the context of asemantically related distractor picture tiger than in thecontext of a semantically unrelated distractor.

In such a framework, the lack of observable semanticeffects in this paradigm might stem from the presence oftwo opposite effects: (a) a facilitatory effect at the con-ceptual level7 (tiger increases the activation of the con-ceptual representation of lion) and, (b) an interfering

7 Semantic facilitation effects in picture–picture tasks havebeen reported by Damian and Bowers (2003). In that study,participants were required to manually categorize objects asman-made or natural. The target objects were presented withdistractor pictures, and the same pairs of objects were used inthe categorization task as were used in the picture–picturenaming experiment. Contrary to picture naming task, where noeffect was observed, a congruency effect was observed in thecategorization task. That is, categorization latencies were fasterwhen target pictures where paired with semantically relateddistractor pictures than with unrelated distractor pictures.

8 Convergent evidence that shallow processing of the distrac-tor picture may lead to semantic facilitation comes from thestudies in which the saliency of the distractor picture ismanipulated. For example, when distractor pictures are pre-sented under difficult perceptual conditions (very briefly ormasked), semantic facilitation effects are observed even whenthe primary task is picture naming (Dell� Acqua & Grainger,1999; La Heij, Heikoop, Akerboom, & Bloem, 2003). Aninteresting observation, also consistent with this idea, is thatwhen the distractor picture is briefly presented and masked, theamount of semantic facilitation is the same for semanticallyrelated distractors as for identical distractors (Dell� Acqua &Grainger, 1999). This observation suggests that under highlydemanding attentional conditions, processing of the distractorpicture is rather shallow.

372 E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377

effect at the lexical level (the lexical node �tiger� competesfor selection with the lexical node �lion�).

This account is admittedly tentative and future re-search needs to evaluate its appropriateness. However,following this account, the presence of phonologicalactivation of distractor pictures is naturally explained,whereas other accounts (of the lack of semantic effects)do not seem appropriate for capturing the PFE. Forexample, Damian and Bowers (2003) assume thatsemantic effects are not present because the semanticrepresentation of the distractor picture does not acti-vate its lexical representation (see also Bloem & LaHeij, 2003). In such a framework, it is a mystery howa distractor picture can activate its phonologicalcontent.

Contrasting results from seemingly similar paradigms

As we argued above, the experimental setting inwhich participants are asked to name a picture (or a col-or) and ignore a distractor object leads to two reliableobservations: (a) a phonological facilitation effect, and(b) a lack of a semantic effect.

However, in a recent series of studies Bloem and LaHeij (2003) found a contrasting pattern of results in aparadigm that involves: (a) conceptually mediated nam-ing, and (b) distractor objects. In their experiments, par-ticipants translated printed words from L2 into L1 whileignoring the presentation of distractor pictures. Bloemand La Heij found semantic facilitation effects and nophonological effects whatsoever. This pattern of resultsis in clear opposition to the one observed when distrac-tor objects are presented in the context of a naming task(e.g., Damian & Bowers, 2003; and our Experiment 1b).What are the reasons for this discrepancy? Why is it thatwhen a paradigm leads to semantic effects it does notlead to phonological effects and vice versa? An answerto these questions requires that we consider the atten-tional processes involved in the different tasks (namingand translation) and how they may interact with theamount of processing carried out over the distractor.

Research from different disciplines shows that theamount of processing that distractors undergo (evenwhen these distractors are supposed to be processedautomatically) is positively correlated with the amountof attentional resources left free by the primary taskconducted by the participant (e.g., Ress, Russel, Frith,& Driver, 1999; Sinnett, Costa, & Soto-Faraco, inpress). Arguably, the attentional load involved in pic-ture naming is smaller than that involved in wordtranslation (e.g., word translation requires a bilingualto have two lexicons activated simultaneously, keepcontrol over them, avoid phonological interferencefrom the to-be-translated word, and perform a cogni-tive task that is much less frequent than naming, seefor example Kroll & Stewart (1994) in which transla-

tion tasks took about 600 ms more than namingtasks). In such a scenario, it is possible that distrac-tors are more fully processed in the naming task thanin the translation task. The differential processing ofdistractor pictures in the two tasks may have impor-tant implications for the presence of semantic andphonological effects.

Shallow processing of a distractor picture in thetranslation task may result in partial activation of itsconceptual representation. Perhaps, in this task, thedistractor only activates certain semantic information(or only structural information) as, for example, cate-gorical membership (e.g., the conceptual informationextracted from the distractor object tiger would beANIMAL). This activation may be enough to primethe conceptual representation of the target (leading toconceptual facilitation), but it might be insufficient toreliably activate the distractor�s lexical representation.As a consequence, lexical competition from the distrac-tor lexical node (e.g., �tiger�) would be minimal. The netresult of this facilitation at the conceptual level andlack of (or very much reduced) lexical interferencewould give rise to the semantic facilitation effect ob-served in translation tasks. In contrast, when the atten-tional demands are lower, as is the case in the picturenaming task, the distractor would be more fully pro-cessed leading to the activation of its conceptual andlexical representations. This situation would lead toboth conceptual facilitation and lexical interference,which will cancel each other out.8

This explanation of the contrasting effects of seman-tically related distractors in different tasks also providesa natural account of the contrasting effects of phonolog-ically related distractors. If semantic effects are restrictedto those experimental conditions in which the distractorpicture is not processed enough to activate its corre-sponding lexical node, then in such conditions oneshould not observe phonological effects. This is becauselexical activation is a pre-requisite for phonological acti-vation. In contrast, those experimental conditions thatallow a more complete processing of the distractor

E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377 373

would lead to its lexical activation and therefore,according to the cascade principle, would lead to itsphonological activation. That is the reason why phono-logical effects are observed in picture naming and notobserved in word translation.9

This attempt to reconcile seemingly contrasting re-sults in different tasks requires future experimentationthat determines the contribution of attentional factorsto the observed effects. Although a detailed discussionof this issue falls outside of the scope of the present re-search, it is appropriate that we briefly discuss howattention may modulate the processes involved in lexicalaccess (e.g., see Ferreira & Pashler, 2002).

Attentional factors regulating cascade processing

The phonological activation of distractor picturesobserved in our study raises the question of to what ex-tent any information that reaches the conceptual systemwill pass activation to the lexical system. As discussedabove, several factors may contribute to whether thisis or is not the case. The fact that we were able to registerphonological activation of to-be-ignored stimuli sug-gests that some of these stimuli activate their lexicaland sublexical representations, regardless of the speak-er�s communicative intention. However, this does notnecessarily imply that any stimulus that reaches thespeaker�s senses is lexically encoded. In fact, our resultsare silent about whether this is the case when individualsare not producing language. Furthermore, even inspeech production contexts, very likely only those stim-uli that reach certain levels of semantic activation wouldbe able to affect the lexical system in some detectablemanner. And the extent to which these stimuli reachthe conceptual system may depend on various factorssuch as the attentional load devoted to other tasks andthe saliency of the irrelevant information. In fact, if

9 This discussion refers to the effects of picture distractors inpicture naming. In other experimental situations in whichdistractor words are presented in the context of picture naming,reliable phonological facilitation, and semantic interferenceeffects are observed. However, this observation does notundermine the arguments developed above. This is because,distractor words and distractor pictures enter into the cognitivesystem from different points (the semantic and the lexicalsystems, respectively). Hence, distractor pictures and distractorwords may be affecting different levels of representation todifferent extents. Also, the attentional resources needed toprocess these two different modalities are very likely different.

the speaker is very focused on the conversation and/orthe task demands a lot of attentional resources (forexample, when speaking in public or in a L2) it is possi-ble that none of the irrelevant information surroundingher is processed enough to affect the lexical system (seefor example the inattentional blindness effect, Mack,2003; Mack & Rock, 1998; Simons, 2000; Simons &Chabris, 1999). Thus, the conditions upon which irrele-vant information can enter into the lexical system mayvary considerably (see Lavie, 1995, for a similar argu-ment on the degree with which distractor stimuli areprocessed in the context of attention theories). What isimportant however for our purposes here is that whenthe conceptual system processes the irrelevant informa-tion to some extent, such activation spreads to subse-quent levels of processing regardless of whether it isselected for lexicalization.

Conclusion

In this article, we put to test different views abouthow information is passed from one level of representa-tion to another during lexical access in speech produc-tion. The basic difference between these views is theextent to which they assume that spreading activationis a governing principle through the production system.We argued that the presence of phonological activationfrom semantically unrelated distractor pictures suggeststhat in the course of speech production, whenever a con-ceptual representation is sufficiently activated, it spreadssome activation to the lexical and phonological levels.That is, these results support the notion that lexicalaccess honors the spreading activation principle at allof its levels of representation. Thus, our observationssupport the notion that activation flows in a cascademanner through the whole production system.

374 E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377

Appendix A

Materials used in Experiment 1a

Target

Distractors

Related

Unrelated

Arpa (harp)

arbol (tree) coche (car) Bate (bat) vaca (cow) mesa (table) Boca (mouth) bota (boot) lapiz (pencil) Camisa (shirt) caballo (horse) pala (spade) Candado (lock) canguro (kangaroo) bota (boot) Casa (house) cama (bed) vaso (glass) Cepillo (brush) cebra (zebra) tortuga (turtle) Collar (necklace) coche (car) platano (banana) Copa (cup) conejo (rabbit) arbol (tree) Corcho (cork) corbata (tie) pipa (pipe) Estrella (star) escoba (broom) puerta (door) Gato (cat) gafas (glasses) corbata (tie) Lazo (bow) lapiz (pencil) plancha (iron) Limon (lemon) libro (book) maleta (suitcase) Luna (moon) lupa (magnifying glass) escoba (broom) Maceta (flowerpot) maleta (suitcase) cebra (zebra) Melon (melon) mesa (table) libro (book) Pato (duck) pala (spade) lupa (magnifying glass) Pina (pineapple) pipa (pipe) conejo (rabbit) Planta (plant) plancha (iron) caballo (horse) Plato (dish) platano (banana) gafas (glasses) Puente (bridge) puerta (door) canguro (kangaroo) Tornillo (screw) tortuga (turtle) vaca (cow) Valla (fence) vaso (glass) cama (bed)

Appendix B

Materials used in Experiment 1b

Target

Distractors

Related

Unrelated

Barco (ship)

avion (plane) mano (hand) Boca (mouth) pierna (leg) armario (wardrobe) Botella (bottle) plato (dish) pierna (leg) Caballo (horse) foca (seal) pantalon (pants) Camisa (shirt) sombrero (hat) serpiente (snake) Coche (car) helicoptero (helicopter) foca (seal) Cuchillo (knife) taza (cup) helicoptero (helicopter) Falda (skirt) corbata (tie) avion (plane) Gato (cat) pez (fish) plato (dish) Helado (icecream) pastel (cake) trombon (trombone) Manzana (apple) uva (grape) sombrero (hat) Martillo (hammer) alicates (pliers) trompeta (trumpet) Mesa (table) armario (wardrobe) uva (grape) Nariz (nose) mano (hand) sarten (frying pan) Pajaro (bird) serpiente (snake) cama (bed) Piano (piano) trompeta (trumpet) taza (cup) Pie (feet) ojo (eye) sofa (sofa)

E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377 375

Appendix B (continued)

Target

Distractors

Related

Unrelated

Platano (banana)

fresa (strawberry) guitarra (guitar) Silla (chair) cama (bed) ojo (eye) Taburete (stool) sofa (sofa) pastel (cake) Tambor (drum) guitarra (guitar) corbata (tie) Vaso (glass) sarten (frying pan) alicates (pliers) Violın (violin) trombon (trombone) fresa (strawberry) Zapato (shoe) pantalon (pants) pez (fish)

Appendix C

Materials used in Experiments 2, 3, and 4

Spanish (Experiments 2, 3, and 4)

Catalan (Experiment 2)

Picture

Color Picture Color

Experimental pictures

Vela (Candle)

verde (green) Espelma verd Ventana (Window) verde (green) Finestra verd Nariz (Nose) naranja (orange) Nas taronja Navaja (Clap knife) naranja (orange) Navaja taronja Roca (Rock) rojo (red) Roca vermell Rodilla (Knee) rojo (red) Genoll vermell Maleta (Suitcase) marron (brown) Maleta marro Mariposa (Butterfly) marron (brown) Papallona marro

Filler pictures

Camion (Truck)

Camio Canon (Canon) Cano Casco (Helmet) Casc Piano (Piano) Piano Sombrero (Hat) Barret Telefono (Phone) Telefon Tenedor (Fork) Forquilla Zapato (Shoe) Sabata

References

Bloem, I., & La Heij, W. (2003). Semantic facilitation andsemantic interference in word translation: Implications formodels of lexical access in language production. Journal ofMemory and Language, 48, 468–488.

Bloem, I., van der Boogaard, S., & La Heij, W. (2004). Semanticfacilitation and semantic interference in languageproduction:Further evidence for the conceptual selectionmodel of lexicalaccess. Journal of Memory and Language, 51, 307–323.

Caramazza,A. (1997).Howmany levels of processing are there inlexical access? Cognitive Neuropsychology, 14, 177–208.

Caramazza, A., & Costa, A. (2000). The semantic interferenceeffect in the picture–word interference paradigm: Does theresponse set matter? Cognition, 75(2), B51–B64.

Costa, A., Miozzo, M., & Caramazza, A. (1999). Lexicalselection in bilinguals: Do words in the bilingual�s twolexicons compete for selection? Journal of Memory and

Language, 41(3), 365–397.Costa, A., Caramazza, A., & Sebastian-Galles, N. (2000). The

cognate facilitation effect: Implications for the models oflexical access. Journal of Experimental Psychology: Learn-

ing, Memory and Cognition, 26, 1283–1296.Costa, A., Santesteban, M, & Cano, A. (2005). On the

facilitatory effects of cognate words in bilingual speechproduction. Brain and Language, 94, 94–103.

Cutting, J. C., & Ferreira, V. S. (1999). Semantic andphonological information flow in the production lexicon.Journal of Experimental Psychology: Learning, Memory, and

Cognition, 25, 318–344.

376 E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377

Damian, M. F. (2000). Semantic negative priming in picturecategorization and naming. Cognition, 76, B45–B55.

Damian, M. F., & Bowers, J. S. (2003). Locus of semanticinterference in picture–word interference tasks. Psychonomic

Bulletin & Review, 10(1), 111–117.Dell, G. S. (1986). A spreading-activation theory of retrieval in

sentence production. Psychological Review, 93, 283–321.Dell, G. S., & O�Seaghdha, P. G. (1991). Mediated and

convergent lexical priming in language production: A com-ment on Levelt et al. (1991). Psychological Review, 98,604–614.

Dell, G. S., Schwartz, M. F., & Martin, N. (1997). Lexicalaccess in aphasic and nonaphasic speakers. PsychologicalReview, 104, 801–838.

Dell� Acqua, R., & Grainger, J. (1999). Unconscious semanticpriming from pictures. Cognition, 73, B1–B15.

Ferreira, V., & Griffin, Z. (2003). Phonological influences onlexical (mis)selection. Psychological Science, 14(1), 86–90.

Ferreira, V. S., & Pashler, H. U. (2002). Central bottleneckinfluences on the processing stages of word production.Journal of Experimental Psychology: Learning, Memory, and

Cognition, 28, 1187–1199.Forster, K. I., & Forster, J. C. (2003). DMDX: A windows

display program with millisecond accuracy. Behavior

Research Methods, Instruments and Computers, 35, 116–124.Glaser, W. R., & Glaser, M. O. (1989). Context effects on

Stroop-like word and picture processing. Journal of Exper-imental Psychology: General, 118, 13–42.

Griffin, Z. M., & Bock, K. (1998). Constraint, word frequency,and the relationship between lexical processing levels inspoken word production. Journal of Memory and Language,

38, 313–338.Harley, T. A. (1993). Phonological activation of semantic

competitors during lexical access in speech production.Language and Cognitive Processes, 8, 291–309.

Humphreys, G. W., Lloyd-Jones, T. J., & Fias, W. (1995).Semantic interference effects on naming using a postcueprocedure: Tapping the links between semantics and pho-nology with pictures and words. Journal of Experimental

Psychology: Learning, Memory, and Cognition, 21, 961–980.Jescheniak, J. D., Hahne, A., & Schriefers, H. (2003).

Information flow in the mental lexicon during speechplanning: Evidence from event-related brain potentials.Cognitive Brain Research, 15, 261–276.

Jescheniak, J. D., & Schriefers, H. J. (1998). Serial versuscascaded processing in lexical access in language produc-tion: Further evidence from the coactivation of near-synonyms. Journal of Experimental Psychology: Learning,

Memory and Cognition, 24(5), 1256–1274.Jescheniak, J. D., Schriefers, H. J., & Hantsch, A. (2001).

Semantic and phonological activation in noun and pronounproduction. Journal of Experimental Psychology: Learning,

Memory and Cognition, 27(4), 1058–1078.Kroll, J. F., & Stewart, E. (1994). Category interference in

translation and picture naming: Evidence for asymmetricconnection between bilingual memory representations.Journal of Memory and Language, 33(2), 149–174.

La Heij, W., Heikoop, K. W., Akerboom, S., & Bloem, I.(2003). Picture naming in picture context: Semantic inter-ference or semantic facilitation? Psychology Science, 45,49–62.

Lavie, N. (1995). Perceptual load as a necessary condition forselective attention. Journal of Experimental Psychology:

Human Perception and Performance, 21, 451–468.Levelt, W. J. M. (1989). Speaking: From intention to articula-

tion. MA: The MIT Press.Levelt, W. J. M. (2001). Spoken word production: A theory of

lexical access. Proceedings of the National Academy of

Sciences of the United States of America, 98(23),13464–13471.

Levelt, W. J. M., Roelofs, A., & Meyer, A. S. (1999). A theoryof lexical access in speech production. Behavioral and Brain

Sciences, 22, 1–75.Levelt, W. J. M., Schriefers, H., Vorberg, D., Meyer, A. S.,

Pechmann, T., & Havinga, J. (1991). The time course oflexical access in speech production: A study of picturenaming. Psychological Review, 98, 122–142.

Lupker, S. J. (1979). The semantic nature of response compe-tition in the picture–word interference task. Memory &

Cognition, 7, 485–495.Mack, A. (2003). Inattentional blindness: Looking without

seeing. Current Directions in Psychological Science, 12(5),180–184.

Mack, A., & Rock, I. (1998). Inattentional blindness. MA: MITPress.

Morsella, E., & Miozzo, M. (2002). Evidence for a cascademodel of lexical access in speech production. Journal of

Experimental Psychology: Learning, Memory, and Cogni-

tion, 28, 555–563.Pallier, C., Dupoux, E., & Jeannin, X. (1997). EXPE: An

expandable programming language for on-line psychologi-cal experiments. Behavior Research Methods, Instruments,

and Computers, 29, 322–327.Peterson, R. R., & Savoy, P. (1998). Lexical selection and

phonological encoding during language production: Evi-dence for cascaded processing. Journal of Experimental

Psychology: Learning, Memory, and Cognition, 24, 539–557.Rapp, B., & Goldrick, M. (2000). Discreteness and interactivity

in spoken word production. Psychological Review, 107,460–499.

Ress, G., Russel, C., Frith, C. D., & Driver, J. (1999).Inattentional blindness versus inattentional amnesia forfixated but ignored words. Science, 286, 2504–2507.

Roelofs, A. (1992). A spreading-activation theory of lemmaretrieval in speaking. Cognition, 42, 107–142.

Rosinski, R. R., Golinkoff, R. M., & Kukish, K. S. (1975).Automatic semantic processing in the picture–word inter-ference task. Child Development, 46, 247–253.

Schmitt, B., Meyer, A., & Levelt, W. J. M. (1999). Lexical accessin the production of pronouns. Cognition, 69, 313–335.

Simons, D. J. (2000). Attentional capture and inattentionalblindness. Trends in Cognitive Sciences, 4, 147–155.

Simons, D. J., & Chabris, C. F. (1999). Gorillas in our midst:Sustained inattentional blindness for dynamic events. Per-ception, 28, 1059–1074.

Sinnett, S., Costa, A., & Soto-Faraco, S. (in press). Inatten-tional blindness across sensory modalities.

Snodgrass, J. G., & Vanderwart, M. (1980). A standardized setof 260 pictures: Norms for name agreement, image agree-ment, familiarity, and visual complexity. Journal of Exper-imental Psychology: Human Learning and Memory, 6,174–215.

E. Navarrete, A. Costa / Journal of Memory and Language 53 (2005) 359–377 377

Starreveld, P. A., & La Heij, W. (1995). Semantic interference,orthographic facilitation and their interaction in namingtasks. Journal of Experimental Psychology: Learning, Mem-

ory, and Cognition, 21, 686–698.Starreveld, P. A., & La Heij, W. (1996). Time-course analysis of

semantic and orthographic context e.ects in picture naming.

Journal of Experimental Psychology: Learning, Memory, and

Cognition, 22, 896–918.Tipper, S. P. (1985). The negative priming effect: Inhibitory

priming by ignored objects. Quarterly Journal of Experi-

mental Psychology: Human Experimental Psychology A, 37,571–590.