processing complex sentences: a cross-linguistic study · pdf fileprocessing complex...

Processing Complex Sentences A Cross-linguisticStudy

Elizabeth BatesUniversity of California San Diego La Jolla CA USA

Antonella Devescovi and Simona DrsquoAmicoUniversity di Roma lsquoLa Sapienzarsquo Rome Italy

The Competition Model is an interactive-activation framework for the studyof sentence processing that is designed to handle quantitative as well asqualitative variations in performance across natural languages Previousstudies within this framework have shown that adult listeners base theirinterpretation of simple sentences on the most validand reliablecues in theirlanguage (eg more use of word order in English and more use of subjectndashverb agreement in Italian) Critics have argued that such effects may reectheuristics that are only applied to simple sentences The present study showsthat these cross-linguistic differences are maintained when participants areasked to interpret complex sentences with an embedded relative clause Acomparison of lsquolsquooff-linersquorsquo (untimed) and lsquolsquoon-linersquorsquo (timed) versions of thesame experiments shows that these effects hold up under time pressure Theon-line versions also provide new information about cross-linguisticdifferences in timing and demands on processing In particular theprocessing costs associated with centre embedding and non-canonical orderare greater in English which may be the price that English listeners have topay for heavy reliance on word order information

INTRODUCTIONThe Competition Model is an interactive-activation framework for thestudy of sentence processing that is designed to handle quantitative as wellas qualitative variations in performance across natural languages

Requests for reprints should be addressed to Elizabeth Bates Center for Research inLanguage 0526 University of California San Diego La Jolla CA 92093-0526 E-mailbatescrlucsdedu

This research was supported by NIDCD grant PHS DC00216 (lsquolsquoCross-linguistic studies inaphasiarsquorsquo) to EB

c 1999 Psychology Press Ltd

LANGUAGE AND COGNITIVE PROCESSES 1999 14 (1) 69ndash123

70 BATES ET AL

(MacWhinney amp Bates 1989) The term lsquolsquocompetitionrsquorsquo reects a centralassumption of the model That different sources of information (ielsquolsquocuesrsquorsquo) converge compete andor conspire to determine the outcome ofsentence processing with different outcomes depending on the relativestrength of cues from one language to another For example the model isdesigned to capture the fact that two languages with the same basic wordorder (eg subjectndashverbndashobject or SVO) can differ markedly in the extentto which listeners rely on word order information to assign semantic rolesin sentence comprehension compared with other sources of informationlike subjectndashverb agreement

In a long series of sentence comprehension studies researchers workingwithin the Competition Model have shown that listeners rely on the mostvalid (ie frequent and reliable) sources of information in their languageMacWhinney and Bates (1989) refer to this effect as lsquolsquocue validityrsquorsquo a termborrowed from Brunswik (1956) English is a language in which word orderis high in cue validity Constituent order is rigidly preserved acrosssentence types and correlates highly with semantic roles (ie lsquolsquowho didwhat to whomrsquorsquo) Italian is in the same typological category (SVO withoutcase marking on nouns) but word order is low in cue validity that isbecause extensive variation of word order is permitted for pragmaticpurposes the correlation between word order and semantic roles isrelatively low Studies of sentence comprehension in these two languageshave shown that English listeners rely primarily on word order to decidelsquolsquowho did the actionrsquorsquo making little use of subjectndashverb agreement orsemantic contrasts Italian listeners are heavily inuenced by subjectndashverbagreement and semantic contrasts but they pay little attention to wordorder The same paradigm has now been used in more than a dozenlanguages (English Italian Spanish French German Dutch Serbo-Croatian Hungarian Japanese Chinese Warlpiri Hebrew ArabicGreek for summaries see Li Bates amp MacWhinney 1993 MacWhinneyamp Bates 1989) It is clear from these studies that many different proles orhierarchies of cue utilization are possible (eg word order gt agreement gtsemantics in English case gt agreement gt semantics gt word order in Serbo-Croatian passive marking gt semantics gt word order gt topic marking inChinese) For any given language the most valid cues also tend to be therst ones used by children (Devescovi et al 1998 Kail 1989) the mostprone to transfer during second-language learning (Hernandez Bates ampAvila 1994 Kilborn amp Ito 1989 Liu Bates amp Li 1992) and the mostresistant to loss following focal brain injury (Bates Wulfeck ampMacWhinney 1991)

In most of these studies listeners are asked to lsquolsquochoose the one who didthe actionrsquorsquo in response to simple sentences or sentence-like stringsconsisting of a transitive verb and two concrete animate or inanimate

PROCESSING COMPLEX SENTENCES 71

nouns These stimuli are constructed to represent orthogonal combinationsof cue types including word order semantic reversibility subject-verbagreement case marking contrastive stress and topic marking Becausefactorial designs are used the stimuli represent all possible competing andconverging combinations of cues to sentence meaning Hence it is possibleto evaluate the relative strength of comparable linguistic forms from onelanguage to another (Massaro 1987) On the other hand this factorialapproach also means that listeners are sometimes faced with a combinationof grammatical and semi-grammatical stimuli Some typical examples fromEnglish would include the following

1 The horse is kicking the cow (NVN AnimatendashAnimate neutralagreement)

2 The dog the cat is chasing (NNV AnimatendashAnimate neutralagreement)

3 Is kissing the boy the girl (VNN AnimatendashAnimate neutralagreement)

4 The ball is pushing the elephant (NVN InanimatendashAnimateneutral agreement)

5 The tiger are chasing the bears (NVN AnimatendashAnimate 2ndnoun agrees)

6 The rocks is hitting the pig (NVN InanimatendashAnimate 2nd nounagrees)

7 Is hitting the rabbit the pencils (VNN AnimatendashInanimate 1stnoun agrees)

8 The boy are pushing the blocks (NVN AnimatendashInanimate 2ndnoun agrees)

To illustrate consider the following ndings for English and Italian onsentences like these and their Italian equivalents replicated in severaldifferent experiments (Bates et al 1982 1984 1987 Devescovi et al1998 Hernandez et al 1994 Liu et al 1992 MacWhinney Bates ampKliegl 1984)

Starting with sentence (1) any theory of sentence processing in Englishor Italian would necessarily predict a SVO interpretation and that isexactly what we nd However our studies have also shown that use ofSVO is quantitatively greater in English than in Italian on semantically andmorphologically reversible items of this kind (averaging 10 more choiceof the rst noun in English from one experiment to another) To explainthis outcome one needs to know something about the relative strength ofcues like SVO in English versus Italianmdasha kind of information that is leftout of many parsing and sentence interpretation theories

72 BATES ET AL

Languages can also vary in the way that listeners respond to non-canonical word orders like the NNV and VNN structures in sentences (2)and (3) In our previous studies using stimuli of this kind we have shownthat English adults are much more likely to choose lsquolsquothe catrsquorsquo in sentence(2) and lsquolsquothe girlrsquorsquo in sentence (3) which are OSV and VOS interpretationsthat do not correspond to any single grammatical structure in thislanguage However this result for English can be explained if we assumethat listeners make use of the partial overlap between semi-grammaticalstimuli and well-formed phrase structure types that do exist in thelanguage For example it is true that subjects are overwhelmingly morelikely to precede the verb in canonical SVO sentences in relative clauseconstructions like lsquolsquoThe boy that the girl kickedrsquorsquo and in left-dislocatedstructures that are permitted in informal speech within some dialects (eglsquolsquoNow this one I likersquorsquo) The conjoint effects of these models could explainthe OSV bias in English Similarly it is also true that objects areoverwhelmingly more likely to follow the verb in canonical SVOsentences in imperative constructions (eg lsquolsquoHit the ball Johnrsquorsquo) and inright-dislocated lsquolsquoafterthoughtrsquorsquo structures that are occasionally observedin informal discourse (eg lsquolsquoMakes a mean apple pie old Gertrude doesrsquorsquo)These structures could explain the VOS bias that English listeners use tointerpret VNN stimuli In contrast with the VOS and OSV patternsobserved in English Italian listeners are close to the random 50baselinein their interpretations of sentences like (2) and (3) presumably becauseOSV SOV VOS and VSO are all permitted in informal Italian discourseunder some pragmatic andor morphological conditions (Beninca 1993Simone 1993) Across several experiments Italians tend to show a slightbias towards SOV and VOS but these trends are not always reliable

In sentences like (4) most English listeners choose lsquolsquothe ballrsquorsquo as theactor which means that SVO word order dominates over semanticcontrasts in this language In addition English listeners usually chooselsquolsquothe tigerrsquorsquo as the actor in sentence (5) which means that SVO word orderalso wins in a competition against subjectndashverb agreement The use ofword order is so strong in English that listeners usually choose lsquolsquothe rocksrsquorsquoin sentence (6) where SVO word order must compete against thecombined forces of animacy and agreement Indeed English listeners eventrust their non-canonical OSV and VOS strategies more than they trustsemantic or morphological information evidenced in the fact that lsquolsquothepencilsrsquorsquo is chosen more often as the actor in sentences like (7) where VOSmust compete against animacy and agreement In contrast Italians showdramatically different patterns of sentence interpretation with equivalentstimuli On sentences like (4) Italians typically choose lsquolsquothe elephantrsquorsquo asthe actor which means that animacy wins in a competition with SVO wordorder This cross-linguistic difference is even more dramatic on items like


(5) where Italians are overwhelmingly more likely to choose lsquolsquothe bearsrsquorsquoas the actor suggesting that subjectndashverb agreement is far more importantthan SVO word order Not surprisingly the combined forces of animacyand agreement defeat word order handily in items like (6) and (7) forItalian listeners Finally on items like (8) where animacy and word ordercompete against agreement English listeners continue to follow their SVObias (choosing lsquolsquothe boyrsquorsquo as actor) but Italian listeners choose lsquolsquotheblocksrsquorsquo as actor showing that subjectndashverb agreement is by far the mostimportant cue within their language

Although a large number of studies have used this method successfullyin many languages the method itself is still controversial (eg Caplan ampHildebrandt 1988 Gibson 1992) Criticisms have focused on three issues(1) Because listeners are asked to make explicit interpretations of agentndashpatient roles (ie lsquolsquowho did what to whomrsquorsquo) the task encourages use ofconscious strategies (2) The presence of ungrammatical stimuli could havea deleterious or unnatural effect on the processes listeners use to interpretany or all of the sentence stimuli (3) Results may reect lsquolsquoshort-cutsrsquorsquo orheuristics that listeners use to interpret simple transitive sentences thesame processes may not hold when listeners are forced to parse andinterpret complex sentence stimuli The rst two criticisms have beenaddressed in several previous studies with arguments that we willsummarise briey here The third criticism will be addressed in thepresent study where we examine the proles of cue utilisation that Englishand Italian speakers use to interpret complex sentence stimuli

The rst criticism can be decomposed into two problems lsquolsquostrategiesrsquorsquoand lsquolsquoconsciousnessrsquorsquo Our answer to the rst problem is that strategies orheuristics are not necessarily a bad thing if the strategies that we elicit inour design generalise to those that are used in everyday life (Kimball1973) In this regard we are comforted by the fact that the performanceproles displayed by native speakers in our cross-linguistic experimentscorrelate highly with independent measures of cue validity and withperformance across different laboratory tasks including grammaticalityjudgement (Devescovi et al 1997 Liu et al 1992 Wulfeck Bates ampCapasso 1991) and word monitoring (Kilborn 1987) With regard to thesupposed problem of consciousness we note that perceptual psychophysicshas made great strides through the use of methods that enlist the consciouscooperation of the perceiver When we engage the subjectrsquos cooperation indeciding lsquolsquowho did the actionrsquorsquo in our cross-linguistic studies we areemloying a similar approach Consciousness per se is not a problem thereal issue is whether the mix of conscious and unconscious processestapped by a given technique generalises to real-time language use undernatural conditionsmdashwhich seems to be the case for the method that wehave used

74 BATES ET AL

The second major criticism revolves around the mix of grammatical andsemi-grammatical stimuli that are used in many of our experiments Toaddress this problem we have carried out several studies in languages inwhich all the stimuli produced by a factorial design are grammaticallyacceptable (eg studies of Serbo-Croatian by Smith amp Mimica 1984 andby Smith amp Bates 1987) In those studies we have found the same effectsof cue validity competition and convergence that emerge in factorialdesigns that use a mixture of grammatical and ungrammatical forms Thispoint was made explicitly by MacWhinney and Pleh (1988) who conductedtwo studies of sentence interpretation in Hungarian in which word orderand animacy were pitted against the presence or absence of nominativeaccusative case marking In one study sentence stimuli without casemarking were all ungrammatical in another the stimuli without casemarking were all grammatical The results were the same in bothexperiments Findings like these have convinced us that our results andconclusions are valid despite the presence of ungrammatical stimuli in afactorial design Indeed we believe that these stimuli can be just asrevealing as the impossible gures and visual illusions that are so oftenused to study competing and converging principles in visual perception(Gregory 1966)

The third criticism pertains to the limitations of a research programbased entirely upon simple transitive sentences Do the results that wehave obtained for short sentence stimuli generalise to the parsing andinterpretation of complex structures Or are they lsquolsquoshort-cutsrsquorsquo that areonly used for simple sentence types (cf Caplan amp Hildebrandt 1988)Thiscriticism has not been addressed systematically in any of our comparativestudies although MacWhinney and Pleh (1988) have conducted a relatedstudy of relative clause interpretation within a single language HungarianIn that study MacWhinney and Pleh took advantage of the fact thatHungarian permits a wide range of word order variations at the level of thematrix clause (SOV SVO OVS etc) As a result they were able to test arange of competing theories of relative clause interpretation that arenecessarily confounded within English For example Sheldon (1974) hasproposed the Parallel Function Hypothesis to explain the fact that object-relatives are more difcult to process than subject-relatives Within asubject-relative (eg lsquolsquoThe girl that pushed the boy opened the doorrsquorsquo) thehead noun lsquolsquogirlrsquorsquo plays the same role within both the matrix and theembedded clause within an object-relative (eg lsquolsquoThe girl that the boypushed opened the doorrsquorsquo) the head noun is the subject in the matrixclause but it takes the object role within the embedded clause Because ofthe opportunities offered by Hungarian MacWhinney and Pleh were ableto show that some of the strategies for parsing relative clauses that havebeen proposed in the English language literature are the by-product of two


or more simple tendencies (eg lsquolsquoTake the speakerrsquos point of viewrsquorsquo) thatare confounded in English but separable in Hungarian

The paper by MacWhinney and Pleh testies to the value of cross-linguistic research to resolve issues that cannot be addressed within a singlelanguage (eg English) However that within-language study does notprovide a direct cross-language test of the cue validity hypothesis forcomplex sentence types In the present study we compare processing oftransitive sentences with a relative clause in English and Italian in order todetermine whether the strong cross-linguistic contrasts that we haveobserved with simple sentence types are preserved when listeners have toparse more complex structures We examine the strategies that are used tointerpret the matrix clause when the rst or second noun is modied by arelative we also look at the strategies that English and Italian listeners useto interpret the relative clause itself

In this regard we need to underscore a paritcularly important andinteresting characteristic of Italian Word order can be varied forpragmatic purposes within both the matrix and the relative clause Withinthe matrix clause all possible orderings of subject verb and object can beobserved in Italian under some combination of pragmatic semanticmorphological andor prosodic conditions These variations are morecommon in informal spoken language (Bazzanella 1994 Duranti amp Ochs1979 Simone 1990) but all of them have been observed in writtendiscourse as well (Devescovi 1986) Word order variation is not unusual inrichly marked languages including case-marked languages like Hungarianas well as languages like Italian that have no case inections but do employa rich set of agreement markers (eg subjectndashverb agreement andagreement between objects and clitic pronouns) However unlike Italianmany languages that permit word order variation within the matrix clausedo not permit the same degree of freedom inside an embedded clauseConsider the following examples

9a The boy saw the woman that is hitting the man (NVN(VN))9b Il ragazzo ha visto la donna che picchia lrsquouomo (NVN(VN))10a The boy saw the woman that the man is hitting (NVN(NV))10b Il ragazzo ha visto la donna che lrsquouomo picchia (NVN(NV))

In English these sentences are absolutely unambiguous (9a) contains asubject-relative in which the woman is the one who does the hitting(NVN(VO)) while (10a) contains an object-relative in which the womanis hit (NVN(SV)) As we shall see these are also the preferred readings inItalian However they are not the only possible readings For morpho-logically ambiguous sentences like (9b) and (10b) two different readingsare possible in Italian In both sentences the woman may be the one who

76 BATES ET AL

does the hitting (so-called subject relatives NVN(VO) or OV)) or the onewho gets hit (so-called object relatives NVN(VS or SV)) The existence oftwo alternative interpretations (albeit at different levels of probability)means that the reliability of word order information is low in Italian inboth the main clause and the relative clause Under these conditions theCompetition Model predicts that Italians will prefer to make use ofagreement information to interpret complex sentences at every level ofthe sentence (ie main andor subordinate clause) Conversely we mayexpect English listeners to ignore agreement contrasts in favour of wordinformation in both the main clause and the relative clause In otherwords the same cross-linguistic differences that we have observed withsimple sentence stimuli will replicate when subjects are required tointerpret much more complex sentence types

We will report results below for a series of studies in the visual modalityThis is in contrast to most studies within the Competition Model whichhave been carried out in the auditory modality Hence in addition to newexplorations with complex sentences we also need to build in a replicationof our results for simple sentences The decision to use visual stimuli wasmade to avoid any confounds that might result from known and unknowneffects of sentence prosody on assignment of agentndashpatient roles (forexamples see MacWhinney et al 1984) The rst two studies wereconducted lsquolsquooff-linersquorsquo with subjects making their response to writtenstimuli in test booklets at their leisure The second two studies wereconducted lsquolsquoon-linersquorsquo with subjects instructed to make their decisions asquickly as possible to sentence stimuli presented one at a time on acomputer terminal The same sentence stimuli were used in both the off-line and on-line studies

GENERAL METHODSStudies of relative clause processing are notoriously difcult to interpretand even more difcult to follow because so many factors are involvedThe factors of interest here include word order variations and agreementvariations in simple sentences (to ensure that our previous ndingsreplicate in the visual modality) and within both the main clause and therelative clause in complex sentences in the latter case we also want toknow if these factors vary depending on the position of the relative clause(ie whether it modies the rst or the second noun in the main clause)These factors would if fully crossed yield a 3 2 2 3 3 2 designfor the complex sentences (with 216 cells) and a 3 3 design for simplesentences (with 9 cells) resulting in a minimum of 1125 trials (assuming nofewer than ve sentences per cell) To get around this combinatorialexplosion we broke the full design into two separate experiments one


focusing on variations in word order in both the main and the relativeclause (with agreement held constant) and another focusing on variationsin agreement in both the main and the relative clause (with order in themain clause held constant in its canonical form) Each of theseexperiments actually contains three sub-experiments that must be analysedseparately simple sentences (to replicate auditory effects in the visualmodality) complex sentences in which participants must interpret themain clause and complex sentences in which participants must interpretthe relative clause Hence the two larger experiments will require a total ofsix separate analyses of variance with language as a between-subjectsfactor Because these experiments were conducted both off-line (untimedin booklet form) and on-line (timed with computer-controlled presenta-tion) with separate subjects participating in the off-line and on-lineversions the study as a whole comprises 12 sub-experiments which mayresult in a large number of effects To reduce the likelihood of false-positives we have set the study-wide alpha level at P lt 001 the onlyeffects discussed are those that reach this conservative level of signicanceTo reduce the readerrsquos workload statistical details for all signicant maineffects and interactions are reported in appendices

Subjects The subjects in all of our experiments are college studentsattending an urban university and they are native speakers of English orItalian The students who participated in Experiment 1 also participated inExperiment 2 (ie the off-line series) in counterbalanced order in a singletest session or in two separate sessions depending on their individualscheduling constraints In contrast Experiments 3 and 4 (ie the on-lineseries) each used a separate group of subjects

Materials To convert our usual lsquolsquowho did itrsquorsquo task for use withcomplex sentences we had to solve a methodological problem If we wantto ask our listeners to decide lsquolsquowho did itrsquorsquo in a complex sentence with threeparticipants and two verbs we have to decide which verb we want them tointerpret on a given item and we need an easy and straightforward way toconvey that instruction Our solution to this problem is called the lsquolsquoWhoDunnitrsquorsquo task All complex sentence stimuli involve two transitive eventsone with a verb referring to a cruel or criminal act (eg lsquolsquoThe waitressshoots the cowboyrsquorsquo) and one with a verb of witness (eg lsquolsquoThe secretarysees the waitressrsquorsquo) Our instructions to the subject are to lsquolsquocatch the badguyrsquorsquo (ie to report which of the three characters in a given sentencecommitted the crime in question) Thus subjects are tacitly required tointerpret the main clause in sentences like lsquolsquoThe waitress that the secretarysees shoots the cowboyrsquorsquo and they are tacitly required to interpret the

78 BATES ET AL

relative clause in sentences like lsquolsquoThe secretary sees the waitress thatshoots the cowboyrsquorsquo For simple sentences only criminal verbs were used

The nouns and verbs used in all the studies that follow are listed inAppendix 1 (for English and Italian) and sample sentences for eachexperiment are listed in Appendix 2 (for English only) For each studylexical items were randomly assigned to sentence conditions To ensurethat item-specic lexical effects could not inuence the results vedifferent versions were created for each experiment each with a differentrandom assignment of words to sentence types

Scoring Since the notion lsquolsquonumber correctrsquorsquo is meaningless in acompetition design of this sort the dependent variable was dened bywhich noun the subjects chose as the actor For sentences in which thecriminal verb is in the main clause a score of 1 was given if the subjectschose the rst noun in that clause as the actor and a score of 0 was given ifthey chose the second noun as the actor If they failed to give any responseat all or if they chose the third noun (in this case the internal noun in therelative clause) they were given a score of 05 A score of 100 wouldmean that the subject always chose the rst noun as actor a score of 0would mean that they always chose the second noun and a score of 50would indicate chance performance For sentences in which the criminalverb is in the relative clause a score of 1 was given if the subjects chose thehead (external) noun as the actor (eg lsquolsquothe waitressrsquorsquo in lsquolsquothe waitress thatshoots the cowboyrsquorsquo) a score of 0 was given if they chose the embedded(internal) noun as the actor (eg lsquolsquothe cowboyrsquorsquo) in lsquolsquothe waitress thatshoots the cowboyrsquorsquo) If they failed to give a response on that item orchose the third noun (in this case the unmodied noun in the matrixclause) they were given a score of 05 For these analyses a score of 100means that the subjects always chose the head noun as actor 0 meansthat they always chose the embedded noun and 50 indicates chanceperformance

EXPERIMENT 1 WORD ORDER VARIATIONSSubjects The subjects for this experiment were 25 American college

students enrolled in undergraduate courses at the University of CaliforniaSan Diego and 25 Italian college students attending the University ofRome lsquolsquoLa Sapienzarsquorsquo All participants were native speakers with minimalbilingual experience (although most have been exposed to a second orthird language in a classroom setting)

Materials and Procedure The 135 sentence stimuli for Experiment 1were printed in a test booklet in randomised order The subjects were


tested individually in a quiet room They were told that each sentencecontained three characters and a single lsquolsquobad actionrsquorsquo and that their jobwas to decide which individual committed the crime They were asked toread each sentence carefully and to indicate for each sentence lsquolsquoWhichcharacter did the bad actionrsquorsquo by underlining or circling the relevant noun

Results and Discussion (see Appendix 3 forstatistical details)

Experiment 1a Word Order in Simple Sentences The main aim of thissub-experiment was to determine whether the cross-linguistic differencesin interpretation of word order observed in previous studies within theauditory modality are preserved with the visual method adopted here The15 simple sentences with no relative clause were analysed in a 2 3 designwith language (English vs Italian) as a between-subjects variable and wordorder (NVN VNN NNV) as a within-subjects variable All results werestatistically reliable including main effects of language and word order anda language word order interaction (Appendix 3) The interaction isillustrated in Fig 1 which shows that English subjects have a strongpreference for SVO OSV and VOS interpretations (for NVN NNV andVNN respectively) while Italian subjects show a somewhat weaker SVO

FIG 1 Language word order interaction for agent choice in simple vs complex sentences(Experiments 1a and 1b)

80 BATES ET AL

bias and very small biases towards SOV and VOS Separate analyses ofvariance for English and Italian showed that these main effects of wordorder are also reliable within each language (also in Appendix 3) Thesendings constitute a visual replication of many previous auditory studiesdemonstrating once again that English speakers lsquolsquotrustrsquorsquo word order morethan their Italian counterparts

Experiment 1b Interpretation of the Main Clause The main aim of thissub-experiment was to determine whether the cross-linguistic differencesin word order use observed in simple sentences are also observed whenone of the nouns in the main clause is modied by a relative This analysiscovered the 60 complex sentences in which the criminal verb appears in themain clause with a verb of witness in the relative clause The data weresubjected to a 2 (English vs Italian 3 (word order within the main clause)

2 (position of the relative clause modifying the rst or second noun) 2(order within the relative clause) mixed analysis of variance with languageserving as a between-subjects factor and the other variables as within-subjects factors

A large number of signicant main effects and interactions were found(see Appendix 3) Based on the cue validity principle of the CompetitionModel the most important effect in Experiment 1b is the language wordorder interaction which tests the prediction that English and Italiansubjects show the same language-specic word order strategies in complexsentences that they show in many studies using simple sentence stimuliThis does indeed appear to be the case Figure 1 plots the data for simpleand complex sentences together (Experiment 1a vs 1b) to facilitatecomparison With added complexity English subjects continue to pursue astrong SVO strategy for NVN strings (86 rst-noun choice vs 90 insimple sentences) a strong VOS strategy for VNN strings (16 rst-nounchoice vs 17 in simple sentences) with a somewhat weaker OSV bias onNNV strings (32 rst-noun choice vs 235on simple sentences) Italiansubjects also show a bias towards SVO although it is slightly atter incomplex sentences (75 rst-noun choice on complex items vs 84 insimple sentences) They also show a slight attening of their SOV bias onNNV (58 rst-noun choice on complex items vs 64 on simplesentences) By contrast their VOS bias is somewhat stronger on complexVNN strings compared with simpler items (33 and 42 respectively)

Overall these results are in line with our predictions When sentences arerendered more complex through the addition of a relative clause Englishand Italian listeners continue to display the same language-specic wordorder strategies that we have observed in this and other experiments usingsimple transitive sentences The primary effect of complexity lies in a slightattening of word order strategies for NVN and NNV strings pushing


them in the direction of the 50 chance baseline The most parsimoniousexplanation for these small differences is that processing is harder incomplex sentence types adding noise to the data

On the other hand the results in Fig 1 also suggest that some wordorder types (especially NNV) appear to be more strongly affected bycomplexity than others The reason for this became clear when weexplored the signicant three-way interaction of language main clauseorder and position of the relative clause (ie modiying the rst vs thesecond noun) To unpack the three-way effect separate analyses ofvariance were conducted for English and Italian When the Italian datawere analysed alone the only reliable nding was a signicant main effectof main clause order (as in Fig 1) In contrast the English-only analysisyielded a signicant two-way interaction between main clause order andrelative clause position illustrated in Fig 2 This gure shows thatperformance on NNV strings is essentially random (ie 48 rst-nounchoice) when the relative clause modies the second noun The reasons forthis total parsing failure are clear when we consider a sample item fromthis set

11 The waitress the cowboy that the baker hears kills (NN(NV)V)

FIG 2 Effects of main clause order and relative clause position on the interpretation of themain clause for English only (Experiment 1b)

82 BATES ET AL

Of all the grammatical and ungrammatical stimuli in this complex designthis is by far the most difcult that is it is the hardest to parse (three nounsare stacked up before the subject encounters a single verb) and the mostdeviant (ie farthest from any well-informed sentence in the Englishlanguage) It seems that we have found the outer limits the point at whichour English subjectsrsquo ability to assimilate semi-grammatical stimuli to well-formed options in the language nally breaks down

Experiment 1c Interpretation of the Relative Clause The aim of thissub-experiment was to determine whether the cross-linguistic differencesthat we have observed for both simple and complex main clauses (ieEnglish speakers rely more on word order than their Italian counterparts)are also observed within the relative clause itself This analysis covers the60 complex sentences in which the criminal verb is located within theembedded clause with the verb of witness in the matrix clause The designand the analysis of variance are identical to Experiment 1b Multiplesignicant effects were found and are summarised in Appendix 3

The most important effect from the point of view of the CompetitionModel is the two-way interaction between language and relative clauseorder illustrated in Fig 3 We pointed out earlier that Italian permitspragmatic word order variation within both the main and the relativeclause Hence relative clause ordering should be a relatively weak cue toagentndashobject relations for Italians The results in Fig 3 show that thisprediction is conrmed Word order biases are in the same direction inboth languages that is VN items are interpreted as subject-relatives (eglsquolsquoThe waitress that shoots the cowboyrsquorsquo) and NV items are interpreted asobject-relatives (eg lsquolsquoThe waitress that the cowboy shootsrsquorsquo) Howeverthis effect is stronger in English On VN items English subjects choose thehead noun as agent 85 of the time compared with 73 in Italian OnNV items English subjects choose the head noun only 15 of the time(ie 85 object-relative interpretations) compared with 355for Italian(ie 645 object-relative interpretations) Hence even in the absence ofconicting morphological or semantic information Italian subjects rely lesson word order to understand a relative clause

The most complex effect in Experiment 1c was a three-way interactionof language main clause order and relative clause order To explore thisinteraction further we conducted separate analyses for English and ItalianThese analyses showed that word order is the only factor inuencingrelative clause interpretation in English (ie relative clause order was theonly signicant effect in the predicted direction) In contrast the Italian-only analysis yielded main effects of word order at both levels of thesentence (main clause and relative clause) together with a signicant two-way interaction Examination of cell means yielded a straightforward


interpretation of these effects A canonical interpretation of the relativeclause is less likely in Italian when a highly marked and difcult word ordertype is used in the main clause On VN items Italians choose the preferredsubject-relative (VO) interpretation 77of the time if the matrix clause isa canonical NVN compared with 71 for VNN and 72 for NNV OnNV items Italians choose the preferred object-relative (SV) interpretationof NV clauses 715 of the time when the matrix clause is a canonicalNVN 70 when the matrix clause is VNN versus a near-random 53 onNNV These subtle but reliable effects testify further to the fragile andpragmatically conditioned nature of word order strategies in Italiancompared with the robust context-independent word order effects that areobserved in English

Summary of Experiment 1We have shown that language-specic word order strategies are observednot only in simple transitive sentences but also in complex sentences inwhich one of those nouns is modied by a relative clause English subjectsshow robust and consistent word order proles in the main clause for allthree word order congurations NVN is interpreted as SVO NNV isinterpreted as OSV and VNN is interpreted as VOS They also display

FIG 3 Effect of relative clause order on the interpretation of the relative in English andItalian (Experiment 1c)

84 BATES ET AL

strong word order biases within the relative clause itself interpreting VNorders as VO (subject relatives where the agent role is assigned to thehead noun) and NV orders as SV (object relatives where the agent role isassigned to the embedded noun) Italians display their own word orderpreferences within the main clause (NVN as SVO NNV as SOV VNN asVOS) with or without a relative and they display reliable effects of wordorder within the relative clause similar in direction to those observed inEnglish However the use of word order is invariably weaker and lessconsistent in Italian than it is in English at both levels of the sentence inline with the cue validity principle of the Competition Model When simpleand complex sentences are compared the effects of complexity are rathersmall restricted primarily to a slight lsquolsquoatteningrsquorsquo of effects in the directionof the 50 random baseline This is particularly true for English items inthe order NNV where a pile-up of nouns prior to the main verb seems toneutralise the OSV strategy that is usually employed for these sentencetypes

EXPERIMENT 2 SUBJECT-VERB AGREEMENTVARIATIONS

In all of our previous experiments Italian subjects made extensive use ofsubjectndashverb agreement to assign sentence roles far more than theirEnglish counterparts When sentences contain a competition betweenagreement and word order agreement information lsquolsquowinsrsquorsquo in Italian andword order lsquolsquowinsrsquorsquo in English Once again however we did not know ifthese cross-linguistic differences would hold up for complex sentencetypes This was the main aim of Experiment 2 As described in the GeneralMethods the stimuli for Experiment 2 were designed to examinecompeting and converging effects of word order and agreement withinsimple sentences (restricted to NVN order only) within the main clause incomplex sentences (also restricted to NVN order only) and within therelative clause itself (both VN and NV)

Subjects The subjects for this experiment were the same 25 Americanand 25 Italian college students who participated in Experiment 1 Theorder of presentation of the experiments was randomised with approxi-mately half of the subjects participating rst in Experiment 1 and halfparticipating rst in Experiment 2

Procedure The 375 sentence stimuli for Experiment 2 were printed ina separate test booklet in randomised order The instructions and testprocedures were identical to those used in Experiment 1



Experiment 2a Agreement in Simple Sentences The main aim of thissub-experiment was to provide a visual replication of the cross-linguisticdifferences in reliance on subjectndashverb agreement that we have observed inauditory studies This part of the experiment contains 15 sentences 5 ineach of three cells (ambiguous agreement agreement in number with therst noun agreement in number with the second noun) analysed within amixed 2 (language) 3 (agreement conditions) analysis of variance Allthree effects were reliable (language agreement and the languageagreement interaction) The interaction is illustrated in Fig 4 which showsthat subjectndashverb agreement is a much stronger cue in Italian than it is inEnglish Recall that all the stimuli in Experiment 2 have the order NVNwithin the main clause English subjects are overwhelmingly more likely tochoose the rst noun as agent on such stimuli 100 on sentences with nomorphological contrast 100 on sentences in which SVO and agreementconverge and 92 on items in which SVO and agreement are incompetition A separate one-way analysis of variance for English subjectsalone showed that the effect of agreement does not even reach signicance

FIG 4 Effects of language and main clause agreement on the interpretation of the mainclause in simple vs complex sentences (Experiments 2a and 2b)

86 BATES ET AL

in this language By comparison Italians choose the rst noun only 87ofthe time on morphologically ambiguous sentences compared with 98 onitems in which SVO and agreement converge and 28 on items in whichSVO and agreement point in other directions (ie agreement wins thecompetition) A separate one-way analysis of variance for Italian aloneyielded a robust main effect of agreement In short this experiment in thevisual modality replicates our previous ndings for English versus Italianusing auditory stimuli (1) canonical word order lsquolsquowinsrsquorsquo in English whilesubjectndashverb agreement lsquolsquowinsrsquorsquo in Italian and (2) canonical word order isweaker in Italian even when there is no competition from other sources

Experiment 2b Interpretation of the Main Clause The main aim of thissub-experiment was to determine whether the cross-linguistic differencesin agreement that we have observed in simple sentences are also seen inthe main clause of complex sentences The design involved a total of 180sentences analysed within a 2 (language) 3 (levels of agreement in themain clause) 3 (levels of agreement in the relative clause) 2 (orderwithin the relative clause) 2 (placement of the relative after the rst orthe second noun) design with language serving as a between-subjectsfactor and all other variables as within-subjects factors A large number ofsignicant main effects and interactions were found (see Appendix 3)including a four-way interaction of language main clause agreementrelative clause agreement and order in the relative clause

The largest effect and the most important for our purposes here is theinteraction between language and subjectndashverb agreement in the mainclause illustrated in Fig 4 (where it is plotted against the analogous effectfor simple sentences from Experiment 2a in order to facilitate compari-son) As Fig 4 shows our predictions were conrmed The same cross-linguistic differences are observed in simple and complex sentence typesthat is SVO word order lsquolsquowinsrsquorsquo in English and subjectndashverb agreementlsquolsquowinsrsquorsquo in Italian The only difference between Experiment 2a andExperiment 2b appears to be a slight attening of results on complex itemsIn other words complexity does exact a small toll but the overall pattern ispreserved This is the most important effect in Experiment 2b but theminor effects deserve exploration To simplify the task of interpreting afour-way interaction we again conducted separate analyses of variance forEnglish and Italian

In the English analysis two effects reached signicance a main effect oforder within the relative clause and an interaction between relative clauseorder and position of the relative within the main clause Cell means forthis interaction break down as follows When the relative clause modiesthe rst noun English subjects choose that noun as the actor in the mainclause 93of the time with a VN relative versus 975with a NV relative


When the relative clause modies the second noun that noun is chosen asthe object of action 98 of the time with a VN relative versus 97 with aNV relative The outliers here (at 93) are sentences of the formS(VO)VO To be sure this is a very small effect but it runs directlycounter to predictions based on the Parallel Function Hypothesis(Sheldon 1974) according to which the lsquolsquosubjecthoodrsquorsquo of the rst nounin a NVN sentence should be enhanced if that noun also plays the subjectrole within the relative clause Why should this be We suspect that ourEnglish subjects are occasionally seduced into a parsing error on sentencesof this type Consider the following item

(12a) The ballerina that sees the policeman shoots the cook(N(VN)VN)

This item may be accidentally parsed as (12b) or (12c)

(12b) The policeman shoots the cook (NVN)(12c) The ballerina sees the policeman that shoots the cook

(NVN(VN))

Whether or not this is the correct explanation we should bear in mind thatthe difference is relatively small (ie 93 vs 98 rst-noun choice) a minorperturbation of an overwhelming tendency for English subjects to applySVO interpretations within the main clause

In the separate analysis of variance for Italian two effects reachedsignicance a large effect of main clause agreement (see Fig 4) and asmall three-way interaction of main clause agreement relative clauseagreement and relative clause order Examination of the cell means for thisthree-way interaction suggests that it is entirely due to sentences that aremorphologically ambiguous in the main clause where Italians do not havetheir favourite cue available (see Table 1) In the absence of their favouritecue they are forced to rely primarily on SVO word order However incontrast to their English counterparts they do not lsquolsquotrustrsquorsquo SVO very muchand can be thrown off by other combinations of information In fact thesmallest effect of SVO (ie 735 choice of the rst noun) occurs inmorphologically ambiguous main clauses that contain a competition withinthe relative clause The relative is in the non-canonical order NV(favouring an object-relative interpretation) but the head noun agreeswith the embedded verb (favouring a subject-relative interpretation)Bearing in mind that subjects are supposed to interpret the main clause inthis part of the experiment in principle it should be possible to ignore therelative clause altogether Indeed the corresponding sentences do notbother English subjects at all but they do bother Italians This may occur

88 BATES ET AL

for two reasons (which are not mutually exclusive) either becauseprocessing load has become so heavy that response moves towards therandom baseline (ie a noise overload) or because the highly markednature of the relative clause persuades Italians to be suspicious ofcanonical interpretations at every level Evidence in favour of the latterinterpretation comes from our previous studies of simple sentences inItalian which showed that Italians are likely to reject an SVOinterpretation in favour of OVS when contrastive stress is used on eithernoun (MacWhinney et al 1984) or when the sentence contains amorphologically ambiguous clitic pronoun (Devescovi 1992) The story iscomplex but the basic intuition is a simple one Italians do not trustcanonical SVO word order and they are willing to suspend default SVOinterpretations under heavily marked or unusual morphological semanticpragmatic andor prosodic conditions

Experiment 2c Interpretation of the Relative Clause The aim of thissub-experiment was to determine whether the cross-linguistic differencesin agreement that we have observed in main clauses (simple or complex)are also observed when participants have to interpret the relative clauseHence the criminal verb is located in the relative clause and the verb ofwitness in the main clause Otherwise the design is identical to that ofExperiment 2b involving a total of 180 sentences analysed within a 2(language) 3 (levels of agreement in the main clause) 3 (levels ofagreement in the relative clause) 2 (order within the relative clause) 2

TABLE 1Effects of Main Clause Agreement Relative Clause Agreement and Relative Clause

Order on Interpretation of the Main Clause for Italians Only (Experiment 2b)

Main Clause Relative Clause Agreement ()Agreement

Ambiguous Head Noun Embedded Noun

AmbiguousVN 880 880 820NV 800 735 925

First nounagreesVN 970 955 910NV 930 925 920

Second nounagreesVN 330 360 385NV 370 350 345


(placement of the relative after the rst or the second noun) design withlanguage serving as a between-subjects factor and all other variables aswithin-subjects factors The analysis of variance yielded a large number ofsignicant effects and these are summarised in Appendix 3

The most important effect for our purposes here is the three-wayinteraction of language relative clause agreement (which should be strongin Italian and weak in English) and relative clause order (which should bestrong in English and weak in Italian) This interaction is illustrated inFig 5 which shows that all of our predictions are roundly conrmed Ininterpreting the relative clause English speakers rely overwhelmingly onword order items in the VN order are interpreted as subject-relatives anditems in the NV order are interpreted as object-relatives Competinginformation from agreement within the clause has virtually no effect inEnglish (see below) In contrast Italians rely overwhelmingly onagreement information If the head noun agrees with the embedded verba subject-relative interpretation is assigned if the embedded noun agreeswith the embedded verb an object interpretation is far more likelyItalians do have the same word order biases as their English counterparts(VN as subject-relative NV as object-relative) but they are relativelyweak showing up primarily on morphologically ambiguous items To

FIG 5 Effects of relative clause order and relative clause agreement on the interpretation ofthe relative in English vs Italian (Experiment 2c)

90 BATES ET AL

explore these effects and the other complex interactions in this data set inmore detail separate analysis of variance were again conducted forEnglish and Italian (Appendix 3)

The results for the English analysis are crystal-clear The only effect toreach signicance at the P lt 001 level was the robust main effect of orderwithin the relative clause Nothing else even approached signicance Thismeans of course that all of the complex interactions listed in Appendix 3are coming from Italian

Turning to the Italian effects the largest and most imortant of the two-way interactions in Italian involves the relationship between order andagreement within the relative clause illustrated in Fig 5 This interactionshows that agreement wins out over default word order across the board inItalian but competition between factors does exact a small toll On VNitems Italians choose the headnounas subject 84of the time if there is nocompetition from morphology (the default subject-relative interpretation)96 when morphology and word order lead to the same interpretation (aconvergence between agreement and the default interpretation) but only38 of the time when morphology and word order are in competition(corresponding to 62 object-relatives the opposite of English interpreta-tions on the corresponding items) On NV items Italians choose the headnoun as subject 29 of the time on morphologically ambiguous items(corresponding to 71 default object-relative interpretations) 11 whenmorphology and word order lead to the same answer (89object-relativesreecting the convergence of agreement and default order) and 90whenagreement and word order are in competition (overwhelming preferencefor a subject-relative assignment the exact opposite of performance byEnglish subjects on analogous items)

The two remaining interactions can be summarised as followsRegardless of agreement conditions within the matrix clause the hugeeffect of relative clause agreement in Italian always runs in the samedirection However this effect is slightly atter on non-canonical items inwhich the second noun in the main clause agrees with its verb (ie itemsthat are likely to receive an OVS interpretation) This may be one moreexample of when markedness (in the main clause) adds to overallcomplexity pushing performance (in the relative clause) slightly closer tothe chance baseline The effect of relative clause agreement is also slightlysmaller (by a few percentage points) when the relative clause is in thesecond position The reason for this latter nding is not at all obviousHowever a look through all the cells in this complex interaction revealedan interesting oddity the attest agreement effect was observed on itemslike the following (from Appendix 2)

(13) The secretary see the cooks that shoot the doctors


These are items in which the verb of witness in the main clause agrees withthe second noun (which should force an OVS interpretation) and wherethe relative clause modies the second noun in the VN order withagreement on the embedded noun (which should force a VS interpreta-tion) In addition to the fact that these items involve a high degree ofcompetition and non-canonical form they have another peculiar char-acteristic If Italians use morphology at all levels to make their assignments(which is usually the case) they arrive at an interpretation in which thecrime victim is busy watching somebody else while he himself is gettingmurdered By reinterpreting these sentences Italians can restore a morepragmatically felicitous situation in which it is the criminal who is watchingout for other individuals who might be on the scene Note however thatsuch pragmatic biases are small and play their greatest role in highlymarked sentences that are particularly difcult to interpret (even thoughthey are grammatically correct in the Italian language)

Summary of Experiment 2The largest effects by far in this experiment are the ones predicted by theCompetition Model English listeners rely exclusively on word order tointerpret the relative clause whereas Italians rely overwhelmingly onagreement information a bias that only breaks down when processing ismaximally overloaded These results complement our ndings in Experi-ment 1 where English word order biases only break down under the worstpossible processing conditions We did nd some minor effects in bothexperiments that are not predicted by the model which we attribute to abreakdown or lsquolsquosofteningrsquorsquo of sentence processing due to combinations ofcomplexity markedness and pragmatic felicity These effects are interest-ing and may be worthy of further investigation but they are very smallcompared with the large cross-linguistic ndings predicted by the model

In view of previous complaints about the use of ungrammaticalsentences in previous studies testing the Competition Model we want toremind readers that all of the sentence types in Experiments 1 and 2 aregrammatical in Italian although some items are so pragmatically markedthat they stretch the denition of grammaticality Many of the sentencetypes in Experiments 1 and 2 are frankly ungrammatical in English but itis clear that English subjects know exactly what to do with them in all but afew cases Despite the difcult and often very odd nature of thesematerials English and Italian subjects respond with consistent andlanguage-specic interpretations that are predicted by cue validity in theirrespective languages

These results suggest to us that the strategies used by English and Italiansubjects in our experiments bear a consistent relationship to the strategies

92 BATES ET AL

used in everyday conversation However this conclusion would bestrengthened if we could show that the same cross-linguistic effects holdup in an lsquolsquoon-linersquorsquo (timed) paradigm First if similar proles of agentassignment are obtained when subjects are tested under time pressurewith relatively fast reaction times then we may conclude with greatercondence that our ndings are related to sentence interpreation strategiesthat listeners use in everyday life Second we should nd cross-linguisticdifferences in reaction time proles reecting convergence andorcompetition between cues that differ in strength in English and ItalianWe would for example expect word order to create larger effects inEnglish conversely we would expect to nd more effects of agreement inItalian This brings us to Experiments 3 and 4 on-line versions of the twocomplex experiments that we have just described In addition to providinga further test of the Competition Model these on-line studies may yieldnew insights into the processing demands associated with complexsentences illuminating the interplay between universal and language-specic factors in sentence processing

EXPERIMENT 3 WORD ORDER VARIATIONSON-LINE

Subjects The subjects for this experiment were 25 American collegestudents enrolled in undergraduate courses at the University of CaliforniaSan Diego and 25 Italian college students attending the University ofRome lsquolsquoLa Sapienzarsquorsquo All participants were native speakers with minimalbilingual experience (although most have been exposed to a second orthird language in a classroom setting) None had participated inExperiments 1 and 2

Materials and Procedure In this on-line version the same 135 sentencestimuli used in Experiment 1 were presented one at a time on a computerscreen at a MacIntosh SI workstation controlled by the Carnegie MellonExperimental Control Shell (ECS) a predecessor of the PsyScope System(Cohen MacWhinney Flatt amp Provost 1993) (For eight subjects in thisexperiment and six in Experiment 4 administration was switched fromECS to the updated PsyScope system Preliminary analyses indicated thatthe two systems did not yield signicant differences in performance dataare pooled across ECS and PsyScope in all subsequent analyses) Subjectswere tested individually in a quiet room The instructions were identical toExperiments 1 and 2 except that subjects were asked to press a button onthe Carnegie Mellon button box as soon as they knew which actor hadcarried out the lsquolsquobad actionrsquorsquo After each button press they were asked tosay the name of the chosen agent aloud this response was recorded


manually by the experimenter If subjects failed to respond within a 5-sectime window the test sentence disappeared from the screen and the nexttrial began

Scoring Scoring of the manually recorded agent assignments (ielsquolsquowho did itrsquorsquo) was identical to the scoring used for the test booklets inExperiments 1 and 2 with one exception Trials in which the subject failedto respond within the 5-sec time window were eliminated prior to analysisof either the decision or the reaction time data and scores for that cellwere averaged to reect performance on the remaining items Such non-responses comprised 137of all responses for English subjects and 130of all responses for Italian subjects in Experiment 3

Results and Discussion (see Appendices 4 and 5for statistical details)Before exploring the reaction time results obtained in this experiment webegan by examining results for agent choice to determine whether theresults obtained off-line in Experiment 1 replicate when subjects areplaced under time pressure All of these analyses described for Experiment1 were repeated using experiment (1 vs 3) as a between-subjects factorAppendix 4 reports only those main effects and interactions that involvedthe factor lsquolsquoexperimentrsquorsquo Despite the complexity of these analyses theresults were strikingly similar in the off-line and on-line versions Therewere a few small but signicant effects involving experiment as a factor Ingeneral these discrepancies involved a sharpening of word order effects inthe on-line version a tendency that was most evident in English Howeverall of these between-experiment ndings were relatively small and none ofthem reected a change in the direction of results compared withExperiment 1 Therefore in the interests of parsimony we will restrictour discussion to reaction time ndings

Experiment 3a Word Order Variations in Simple Sentences The aim ofthis sub-experiment was to determine whether cross-linguistic differencesin reliance on word order are reected in reaction time A 2 (English vsItalian) 3 (word order in the main clause) analysis of variance yielded asignicant main effect of word order and a signicant interaction betweenlanguage and word order (see Appendix 5 for statistical details) The maineffect of language was not reliable indicating that English and Italiansubjects take approximately the same amount of time to read thesesentences (ie a mean of 2439 msec less then 25 sec from stimulus onset)These relatively fast reaction times (less then 600 msec per word) suggestto us that subjects are applying familiar strategies borrowed from those

94 BATES ET AL

that are used in everyday life The interaction between language and wordorder is illustrated in Fig 6 which shows that English subjects respondedfaster than Italians on NVN (1895 vs 2092 msec) and VNN ((2310 vs2661 msec) and slower on NNV (3024 vs 2664 msec) Hence we mayconclude that word order has a larger effect on the processing of simplesentences in English than Italian an effect that is compatible with cross-linguistic differences in agent choice

Experiment 3b Interpretation of the Main Clause The main aim of thissub-experiment was to determine whether cross-linguistic differences inreliance on word order are reected in reaction times for complex as wellas simple sentences In addition we hoped that this analysis would yieldnew information about the specic processing costs associated withsentence embedding in canonical versus non-canonical word order typesThe materials and design were identical to those of Experiment 1bSignicant ndings (listed in Appendix 5) included main effects of allvariables a three-way interaction of language main clause and relativeclause order and a three-way interaction of language main clause orderand relative clause position Although this is an intricate nest of ndings

FIG 6 Reaction times for the interpretation of the main clause in English and Italian insimple vs embedded sentences (Experiments 3a and 3b)


all of them are quite sensible reecting variations in processing load thatfollow from the strategies used in each language

First the main effect of order and its intereaction with language aresimilar to the results that we have already encountered for simplesentences Canonical NVN is the fastest order in both languages(averaging 3341 msec in English and 3423 msec in Italian) followed byVNN (4172 msec in English and 3950 msec in Italian) with the slowestRTs observed on NNV items (5018 msec in English and 4358 msec inItalian) Of course these reaction times are considerably slower than theaverages obtained with simple sentences reecting the additional timerequired to read and interpret sentences with an embedded clause Thevery sharp increase in RTs for NNV reects the fact that these items sufferthe most from centre embedding (ie N(rel)NV and NN(rel)V are bothcentre-embedded items) This situation weights most heavily on Englishsubjects who are struggling to apply their non-canonical OSV strategy tothe main clause in these difcult embedded structures It has less of aneffect on Italian subjects who have very weak biases on both VNN andNNV and treat the two indifferently with or without an embedded clauseThe language main clause order interaction for Experiment 3b is alsopresented in Fig 6 to facilitate comparison of performance in simpleversus complex sentences

Second the effect of relative clause order on interpretation of mainclauses reects the additional processing load that arises with the presenceof a non-canonical relative (even though subjects do not have to interpretthe relative in this section of the experiment) Collapsed over language andother sentence effects reaction times to interpret the main clause averaged3898 msec when the relative clause was a canonical VN compared with4190 msec when the relative as a non-canonical NV The interactionbetween language and relative clause order shows us that this effect ofcanonicity is much worse for English subjects In fact separate analyses ofvariance within each language showed that the clause order effect washighly reliable for English but not for Italian The three-way interaction oflanguage main clause order and relative clause order is illustrated in Fig 7which shows that English subjects nd it especially difcult to interpretitems in which the main clause is in the order NNV and the embeddedclause in the order NV a situation of maximal embedding with non-canonical word order at both levels of the sentence (eg lsquolsquoThe secretary thecowboy that the policeman shoots seesrsquorsquo) These conditions are apparentlyfar less taxing for Italians who come to the task with less pronounced wordorder biases at either level

Finally the position effects also reect the problem posed by centreembedding (ie interruption of the main clause) This problem isparticularly serious for English subjects who are trying hard to use their

96 BATES ET AL

word order strategies in every case Overall the main effect of relativeclause position reects slower RTs when the relative modies the rstnoun in the main clause (mean = 4140 msec) and faster RTs when themain clause modies the second noun (mean = 3946 msec) However thisposition effect is accounted for entirely by VNN and NVN strings whereearly placement of the clause results in a centre embedding whereas lateplacement does not On NNV strings where centre embedding occursregardless of relative noun position faster results are obtained with earlyplacement In fact NNV items present the most severe problems of centreembedding overall a situation that is apparently exacerbated when therelative modies the second noun (giving sentences of the form NN(rel)V)This is true for both languages but it is especially true for English as wecan see from the three-way interaction of language main clause order andrelative clause position illustrated in Fig 8

To summarise the ndings for this section we again nd signicantcross-linguistic differences in the direction and magnitude of word ordereffects in the main clause similar to the effects we have observed in manystudies using simple transitive sentences with no embedding Reactiontimes reect larger effects of word order for native speakers of English

FIG 7 Reaction times for the interpretation of the main clause as a function of languageorder in the main clause and order in the relative (Experiment 3b)


who also pay a greater price than Italians do for centre embedding andnon-canonical order

Experiment 3c Interpretation of the Relative Clause The main aim ofthis sub-experiment was to determine whether the cross-linguisticdifferences in RT to word order variations reported for simple andcomplex sentences are also observed within the relative clause itself Thematerials and design are identical to those of Experiment 1c covering the60 complex sentences in which the criminal verb is located within theembedded clause with the verb of witness in the matrix clause Theanalysis of variance yielded a large number of signicant effects which aresummarised in Appendix 5

Again although this is an intricate set of statistical ndings theirinterpretation is relatively straightforward Most of the effects in thisanalysis are in the same directions for English and Italian and most ofthem reect three general factors processing is slower in the presence ofnon-canonical word orders at either level (relative or main clause)processing is slower with heavily embedded items and interpretations ofthe relative clause tend to be faster when that clause occurs early in thesentence (modifying the rst noun) As we shall see the second factor (a

FIG 8 Reaction times for the interpretation of the main clause as a function of main clauseorder and relative clause position in English vs Italian (Experiment 3b)

98 BATES ET AL

disadvantage for centre embedding) and the third factor (an advantage ofearly placement of the relative clause within the sentence ie an lsquolsquoearlybirdrsquorsquo effect) work against each other in some sentence types

First the effect of main clause order (signicant in both languages)means that subjects nd it easier to interpret the relative clause when thematrix clause is a canonical NVN (mean RT = 3516 msec for NVN4123 msec for VNN and 4134 msec for NNV) This is a straightforwardeffect of processing load Similarly the effect of relative clause order (alsosignicant in both languages) means that interpretation of the relativeclause is faster when that clause is a canonical VN (mean= 3742 msec) andslower when it is a non-canonical NV (mean = 4107 msec) This can alsobe interpreted as an effect of processing load andor an effect of frequencyof clause types (VN is more frequent than NV in both languages) Theinteraction between main clause and relative clause order is illustrated inFig 9 which shows that the fastest reaction times are observed when bothlevels of the sentence are canonical (NVN sentences with VN relatives) Inseparate analyses this interaction only reached signicance in English(Appendix 5)

The effects involving relative clause position are a bit more subtle butthey derive primarily from the conict between availability (an advantage

FIG 9 Reaction times for the interpretation of the relative clause as a function of wordorder in both the main and relative clause (Experiment 3c)


for early placement) and structural difculty (especially degree ofembeddedness) Overall a relative clause that occurs early in the sentence(modifying the rst noun mean = 3796 msec) can be interpreted soonerthan a relative clause that occurs relatively late (modifying the secondnoun mean = 4053 msec) However in some sentence types the lsquolsquoearlybirdrsquorsquo also produces a more thickly embedded sentence a fact whichcancels out any advantages that early placement might otherwise affordThis is evident in Fig 10 which displays the only effect involving languagein this analysis a complex three-way interaction of language main clauseorder and relative clause position On NVN items which are easy toprocess early placement of the relative provides a substantial advantage inEnglish despite the fact that early placement also results in centreembedding (ie N(rel)VN is easier than NVN(rel)) On VNN items whichare more difcult to process early placement slows things down in bothlanguages (but more so in English) presumably because early placementresults in embedding (VN(rel)N) but late placement does not (VNN(rel))On NNV items which are by far the hardest to interpret early placementoffers a substantial advantage (especially in English) and late placementoffers a substantial disadvantage (again especially in English) Presum-ably this is the case because embedding occurs in NNV no matter where

FIG 10 Effects of main clause order and relative clause position on time to interpret therelative clause in English vs Italian (Experiment 3c)

100 BATES ET AL

we place the relative (N(rel)NV and NN(rel)V are both centre-embeddedstructures) leaving the advantageof early placement to operate on its ownBoth these factors have a larger effect in English

Finally the three-way interaction of main clause order relative clauseorder and relative clause placement is illustrated in Fig 11 This result(which does not interact with language) also reects the interactionbetween early availability (a good thing) and embeddedness (a bad thing)as they relate to canonicity (ie VN relatives are easier than NV relatives)

To illustrate this last point note that the easiest relative clauses amongthe 12 conditions illustrated in Fig 11 are canonical VN forms that occur ina canonical NVN (in either position) and canonical VN forms that occurvery early right after the rst noun in an NNV (resulting in sentences likelsquolsquoThe secretary that shot the cowboy the policeman sawrsquorsquo) Although thislast structure is non-canonical and heavily embedded at the level of themain clause it may be easy when subjects have to interpret the relativebecause they can stop paying attention to the rest of the sentence afterthey have processed a well-formed opening fragment (ie lsquolsquoThe secretarythat shot the cowboy rsquorsquo)

In the same vein the hardest relative clauses among the 12 conditions inFig 11 are those that involve lethal combinations of non-canonical relative

FIG 11 Reaction times for the interpretation of the relative clause as a function of mainclause order relative clause order and relative clause position (Experiment 3c)


clause order (NV) in a non-canonical matrix (NNV or VNN) with thickcentre embedding (eg the three worst are NN(VN)V NN(NV)V andVN(NV)V in that order) In other words horrible sentences take a longtime to interpret in both languages although the price appears to be moresevere in English

Summary of Experiment 3Reaction time ndings in Experiment 3 are intricate but they boil down tothe interaction of some relatively straightforward principles (1) reactiontimes are faster for canonical word order types in both the main clause andthe relative clause (2) centre embeddings take longer to interpret and (3)interpretation of the relative clause is faster (all other things being equal)when that clause is encountered early in the sentence These factorsoperate in both languages but they exact a great reaction time cost inEnglish leading to the speculation that embedding costs may be higher in alanguage that relies heavily on word order information (see Hawkins 1994for an extensive discussion of processing costs and linguistic typology)

This brings us to our nal comparison an on-line version of Experiment2 investigating the effects of variations in subjectndashverb agreement at bothlevels of the sentence in relation to relative clause order and relativeclause placement In this experiment (as in Experiment 2) order in themain clause is held constant in the canonical NVN Again we expect tond larger effects of agreement in Italian and larger effects of relativeclause order and relative clause placement in English in both agentassignment (noun choice) and reaction times

EXPERIMENT 4 SUBJECTplusmn VERB AGREEMENTVARIATIONS

Subjects The subjects for this experiment were 23 American and 23Italian college students meeting all the criteria outlined earlier None ofthese subjects had participated in Experiments 1 2 or 3 For analysescomparing Experiments 2 and 4 the last two of the 25 subjects inExperiment 2 were dropped to equalise sample size

Procedure The stimuli were identical to those used in Experiment 2The instructions test procedures and scoring criteria were identical tothose used in Experiment 3

Scoring Trials on which subjects failed to respond within the 5-sectime window were removed from the analysis with mean agent assignmentscores and mean reaction times based on the remaining trials within each

102 BATES ET AL

cell A total of 027of all trials for English subjects and 064of all trialsfor Italian subjects were removed owing to non-response

Results and Discussion (see Appendices 4 and 5for statistical details)Before exploring the reaction time results obtained in this experiment webegan by examining the results for agent choice to determine whether theresults obtained off-line in Experiment 2 replicate when subjects areplaced under time pressure All of the analyses described for Experiment 2were repeated using experiment (2 vs 4) as a between-subjects factor (seeAppendix 4) The results were identical for Experiments 2a and 4a and forExperiments 2b and 4b (ie there were no main effects of interactionsinvolving the factor lsquolsquoexperimentrsquorsquo) Results for the third part of theexperiment (interpretation of the relative clause Experiment 2c vsExperiment 4c) were also similar in direction and magnitude althoughthere were a few small effects involving experiment as a factor All of theseeffects were signicant in Italian but not in English Examination of cellmeans led us to the culprit The worst cell occurs in the Italian on-linecondition where an NV clause (non-canonical) occupies rst position(maximal embedding) with ambiguous relative clause agreement (so thatItalians cannot rely on their favourite cue) Italians are driven to randomperformance in this cell (504 object-relative interpretations comparedwith 75 object-relative interpretations in the equivalent cell off-line)Interestingly English subjects reach an object-relative interpretation935 of the time in the same on-line cell Hence the processing costsexperienced by English and Italian speakers diverge in this particularcondition suggesting that English listeners are resistant (because they arenot trying to use agreement information) while Italians are vulnerable(because they are trying very hard to use agreement information under theworst possible conditions) In all other respects the off-line and on-lineversions yield comparable results for agent choice Therefore in theinterests of parsimony we will restrict our discussion of Experiment 4 toreaction time ndings

Experiment 4a Agreement in Simple Sentences The main aim of thissub-experiment was to determine whether cross-linguistic differences inthe use of subjectndashverb agreement are also reected in reaction times forsimple sentences in the visual modality Reaction times for the 15sentences in this sub-experiment were analysed in a 2 3 design withlanguage treated as a between-subjects factor and agreement as a within-subjects factor (see Appendix 5 for details) The main effect of languagewas large and reliable reecting much faster RTs in English (mean =


1416 msec) than Italian (mean = 2170 msec) Note also that the overallmean RTs in this experiment (1793 msec) are 761 msec faster than themean RTs observed for simple sentences under the word order variationsmanipulated in Experiment 3a (2440 msec) Both these ndings are due tothe fact that all sentences in this section are in the canonical NVN order asituation that leads to substantially faster reaction times overall butespecially in English The main effect of agreement did not reachsignicance by the P lt 001 standard nor did we nd the expectedinteraction between language and agreement This contrasts with otherexperiments in our laboratories in which Italians were faster than Englishsubjects with converging agreement (ie agreement with the rst noun onNVN) and slower with competing agreement (ie agreement with thesecond noun on NVN) As we shall see below the predicted interactiondoes appear in main clauses modied by a relative (Experiment 4b)

To summarise the results for simple sentences Italians show muchstronger effects of agreement variation than their English counterparts inoff-line and on-line agent choice Contrary to our expectations thisinteraction does not show up in reaction times However the RT data doshow that English listeners are much faster overall to interpret thesecanonical NVN strings

Experiment 4b Interpretation of the Main Clause The main aim of thissub-experiment was to uncover cross-linguistic differences in the RTproles associated with differential use of subjectndashverb agreement withinthe main clause in complex sentences The materials and design wereidentical to those in Experiment 2b with 180 complex items in which thecriminal verb is located in the main clause This RT analysis yielded a largenumber of signicant effects (see Appendix 5) including a signicant four-way interaction of language with agreement order and position within therelative clause

The main effect of language reects much faster reaction times inEnglish (mean = 2444 msec) than Italian (mean = 3433 msec) Bearing inmind that all the main clauses in Experiment 4b are in the canonical NVNorder this result reects the basic ndings for agent choice describedearlier That is English listeners apply their SVO strategies across theboard despite the various manipulations that we have imposed whileItalians vacillate between SVO and OVS in interaction with other factorsAs a result English listeners are much more efcient in this part of thestudy while Italians pay a price for their willingness to accept OVSinterpretations

The interaction between language and main clause agreement conrmsour prediction that Italians are inuenced more by a competition betweenmain clause order and main clause agreement For Italians reaction times

104 BATES ET AL

averaged 3157 msec with rst-noun agreement 3087 msec for ambiguousitems and 4056 msec for competition items (which are usually interpretedas a non-canonical OVS) For Americans reaction times averaged 2361msec with rst-noun agreement 2394 msec for ambiguous items and 2576msec for competition items (which are interpreted as a canonical SVO)The competition effect proved to be reliable in both languages when thedata were analysed separately so this language difference is simply amatter of degree

All of the remaining main effects and interactions can be understood interms of variations in processing load Factors that slow down interpreta-tion of the matrix clause include non-canonical agreement (with the secondnoun in the main clause with the embedded noun in the relative clause)non-canonical word order (NV in the relative) and centre embedding(when the relative modies the rst noun in these NVN strings) Thesefactors apply to some extent in both language groups However ourndings for agent assignment clearly indicate that English speakers lsquolsquotrustrsquorsquoword order and Italian speakers lsquolsquotrustrsquorsquo agreement at both levels of thesentence Hence in this sub-experiment we should expect bigger effects ofrelative clause order and embedding on reaction times in English andbigger effects of main and relative clause agreement on reaction times inItalian The results are largely in line with these predictions

First the main effect of relative clause order reects faster RTs in thepresence of a canonical VN relative (mean = 2849 msec) and slower RTswhen the main clause is modied by a non-canonical NV relative (mean =

3028 msec) This variable does not interact with language which meansthat both language groups slow down in the presence of an NV relative(note that the main effect of relative clause order just misses signicanceP lt 0012 when English is analysed separately see Appendix 5)

Second the main effect of relative clause position reects faster RTswhen the relative modies the second noun (no centre embedding mean =

2384 msec) and slower RTs when the relative modies the rst noun (withcentre embedding mean = 3492 msec) This embedding effect also holdsin both languages but the signicant language position interaction tellsus that the cost of centre embedding is greater in English (mean RT inposition 1 = 3179 msec vs 1710 msec in position 2) than Italian (mean RTin position 1 = 3807 msec vs 3059 msec in position 2)

Finally there was no signicant main effect for relative clauseagreement but this variable did interact with a number of other factorsincluding language For the sake of economy we will restrict ourdiscussion to the four-way interaction because this higher-order interac-tion subsumes most of the other effects As before we began ourexploration of this complex interaction by conducting separate analyses inEnglish and Italian


Figure 12a illustrates the relationships among relative clause agreementorder and position in Italian This three-way interaction was not reliablefor Italians alone Hence the pattern for Italians in Fig 12a actually reectsthree separate results (1) a main effect of relative clause order (withslower RTs in the presence of an NV relative) (2) a large main effect ofposition (with slower RTs with the RT in position 1) and (3) a two-wayinteraction of relative clause agreement and relative clause order Theinteraction (see Fig 12a) shows that the advantage of canonical order(VN) within the relative clause is wiped out in Italian when agreementruns in the opposite direction

The corresponding results for English are illustrated in Fig 12b Aseparate analysis of the English data shows that the full three-wayinteraction is statistically reliable The most obvious feature in Fig 12b isthe massive effect of embedding with RTs in centre-embedded sentencesthat are more than 1000 msec longer than RTs on sentences that end in arelative clause It is also clear from Fig 12b that the RT differencebetween lsquolsquoeasyrsquorsquo VN and lsquolsquohardrsquorsquo NV relatives is relatively small whichmeans that the distracting effect of a non-canonical relative is not astroublesome for the English group in this part of the experiment For ourEnglish subjects the most troublesome cell in the design contains itemswith the relative clause in the rst position (embedded sentences) in NVorder (a non-canonical form) where the verb inside the relative clauseagrees with the head noun (in direct competition with an NV interpreta-tion) Hence English listeners are not entirely immune to the distractingeffects of relative clause agreement but such effects are only observedunder the worst possible conditions of embedding and non-canonicity

To summarise the results for interpretation of the main clause there arelarger effects of word order and embedding in English and larger effects ofagreement in Italian in accord with the predictions of the CompetitionModel However processing costs are in the same general direction forboth groups (ie slower RTs with centre embedding and in the presence ofa non-canonical relative clause)

Experiment 4c Interpretation of the Relative Clause The main aim ofthis nal sub-experiment was to assess cross-linguistic differences in theprocessing costs associated with variations in agreement inside the relativeclause In particular we predicted that agreement effects within therelative would be greater for Italians The materials and design wereidentical to those in Experiment 2c Once again there were a large numberof signicant effects (see Appendix 5) including a four-way interaction oflanguage with all three relative clause factors (agreement and order withinthe clause and position of the clause within the sentence)

(b)

(a)

106

FIG12 (a) Effects of relative clause agreement order and position on time to interpret themain clause in Italian only (Experiment 4b) (b) Effects of relative clause agreement orderand position on time to interpret the main clause in English only (Experiment 4b)


All of these effects reect a combination of cross-linguistic differences(ie relative clause agreement is more important in Italian relative clauseorder and position are more important in English) and variations inprocessing load that apply in both languages The processing facts (whichmay be universal) are the same ones described earlier (1) reaction timesare slower in the presence of non-canonical forms (when the verb in themain clause agrees with the second noun when either word order oragreement in the relative clause favour an object-relative interpretation)and (2) reaction times are slower in the presence of embedding although(3) early placement of the relative can provide an RT advantage (an lsquolsquoearlybirdrsquorsquo strategy)

First the main effect of agreement within the main clause was in linewith our predictions reecting slower performance with agreement on thesecond noun (a competition between default SVO and agreement-inducedOVS) As expected separate analyses for each language showed that thiseffect only reached signicance in Italian In other words because Englishspeakers do not care very much about agreement a violation of agreementin the main clause does not delay their interpretations of the relative Incontrast the same effect increases processing load for Italians and slowsthem down across the board

Second the main effect of agreement within the relative clause reectsslower performance when agreement supports an object-relative inter-pretation (mean = 3289 msec) compared with roughly similar RTs formorphologically ambiguous relatives (mean = 3051 msec) and for items inwhich agreement fosters a subject-relative interpretation (mean = 3079msec) In accord with our predictions separate analyses within eachlanguage showed that this agreement manipulation does not reachsignicance in English but is highly reliable in Italian Hence it is theItalian group that is most affected by agreement manipulations at bothlevels of the sentence

Third the main effect of relative clause order reects faster performanceon canonical VN clauses (mean = 2940 msec) and slower performance onnon-canonical NV clauses (mean = 3339 msec) In this case there was notwo-way interaction with language indicating that relative clause order isimportant for both groups However there was a reliable three-wayinteraction of language with relative clause agreement and relative clauseorder illustrated in Fig 13 Several aspects of our results are evident inFig 13 including the much greater reaction time advantage for Englishsubjects who rely faithfully on word order and pay little attention toagreement It is also clear from Fig 13 that the three-way interaction iscoming from Italian where there is a crossover interaction between orderand agreement within the relative clause Italians are slower on NV clauseswhen agreement is ambiguous or when agreement and NV order are in

108 BATES ET AL

competition but they are faster on NV clauses when agreement and orderconverge These reaction time results mirror our ndings for agent choiceboth on-line and off-line and they are also similar to the subtle three-wayinteractions observed for interpretation of the main clause in Experiment4b

The remaining effects all involve relative clause position and it is herethat we begin to see some language-specic processing costs in EnglishThe main effect of position is the same in both languages an lsquolsquoearly birdrsquorsquoeffect with faster RTs for early relatives (mean = 2974 msec) and slowerRTs for later relatives (mean = 3305 msec) However the advantages ofearly placement may be counteracted in some cases by the costs associatedwith centre embedding In fact separate analyses of English and Italianreveal that all interactions of relative clause position with the other factorsare coming from the English group This includes the interaction betweenrelative clause order and position and the three-way interaction ofposition order and agreement within the relative clause The three-wayinteraction for English alone is presented in Fig 14 which illustratesseveral interesting facts First there is no difference between canonical VNand non-canonical NV when these clauses occur late in the sentence (ie

FIG 13 Effects of order and agreement within the relative clause on time to interpret therelative clause (Experiment 4c)


with no centre embedding) nor can we nd any evidence that relativeclause agreement makes a difference when the clause occurs in sentence-nal position outside of the matrix sentence Second English speakersrespond very quickly and efciently to lsquolsquoearly birdrsquorsquo relatives that modifythe rst noun but only if the relative is a canonical VN In this case wealso nd some facilitation from converging agreement These ndingsmake sense if we remember that these items lend themselves to earlyinterpretation with processing interrupted half way through the sentence(eg lsquolsquoThe secretary that shoots the cowboys rsquorsquo) This is less true for NVclauses in rst position particularly when there is competition betweenorder and agreement (eg lsquolsquoThe secretaries that the cowboy shoot rsquorsquo)In these cases the advantage of early placement is apparently overcome bythe disadvantages of embedding and non-canonical form

In general the results for this section support the view that word order isthe major factor inuencing interpretation of the relative clause in Englishalthough English reaction times are also affected by degree of embeddingand the presence of non-canonical forms Subjectndashverb agreement is themajor determinant of relative clause interpretation in Italian althoughnon-canonical word order does exact a cost

FIG 14 Effects of order agreement and position of the relative clause on time to interpretthe relative clause for English only (Experiment 4c)

110 BATES ET AL

Summary of Experiment 4We have uncovered some extraordinarily complex effects in thisexperiment but the major ndings are in the directions predicted by theCompetition Model English subjects show a marked RT advantageoverall (due to the canonical order of the matrix clause) but they alsodisplay greater effects of centre embedding interacting with canonicalform in the relative clause Italian subjects show pronounced effects ofsubjectndashverb agreement at both levels of the sentence evidenced in largedelays when canonical word order and subject-verb agreement are incompetition in either the main clause or the relative clause At the sametime Italians show relatively few effects of embedding in their reactiontime data Hence our results suggest that the costs associated with centreembedding are greater in languages that rely heavily on word order andlighter (although still present) in languages that rely more on morpholo-gical information We may speculate that the costs of centre embeddingare lower in a richly inected language because listeners can sometimesrely on lsquolsquolocalrsquorsquo solutions to agent interpretation while English listenershave to keep the topology of the sentence in mind in order to reach aninterpretation

CONCLUSIONSThe main aim in this study was to determine whether (and to what extent)the cross-linguistic differences in sentence interpretation that we haveobserved with simple sentences also generalise to complex sentences withan embedded clause After this long journey we can say with somecondence that the answer is lsquolsquoyesrsquorsquo Native speakers of English relyprimarily on word order information in simple sentences in the mainclause of sentences with a relative clause and inside the relative clauseitself This tendency is evident in their sentence interpretations both on-line and off-line and in the speed with which solutions are reached in theon-line versions of each experiment Native speakers of Italian relyprimarily on subjectndashverb agreement in both the main clause and therelative clause despite variations in word order canonicity and degree ofcentre embedding This tendency is also evident on-line and off-line inagent choice and in reaction times

In both languages these tendencies do break down under somestructural conditions For example English speakers fall towards chancewhen they have to interpret horric centre embeddings like lsquolsquoThe secretarythe cowboy that the policeman shoots seesrsquorsquo Italian speakers also falltowards chance when they are forced to interpret non-canonical centre-embedded structures with no morphological contrasts Hence there are


clear limits on processing strategies that characterise and differentiateperformance in these languages

Between the two extremes of language-specic performance andrandom behaviour the reaction time results of Experiments 3 and 4provide some insights into universal processing costs and their language-specic instantiations For both language groups processing is moredifcult (and reaction times are slowed accordingly) in the presence ofnon-canonical structures at either level of the sentence This includes non-canonical word orders (NNV and VNN in the main clause NV in therelative) and forms of agreement that compete with canonical word order(agreement on the second noun in an NVN main clause agreement withthe embedded noun in a VN relative or with the head noun in an NVrelative) In both languages centre embedding can slow the interpretationof a main clause although the magnitude of this embedding cost dependson other factors Early placement can hasten the interpretation of arelative clause (ie the lsquolsquoearly birdrsquorsquo effect) but this factor also interactswith other structural facts (especially the interplay between earlyplacement and centre embedding)

Although these processing costs are observed in both languages (andmight be put forward as candidates for lsquolsquoprocessing universalsrsquorsquo) there arealso robust cross-linguistic differences in the price that must be paidEnglish listeners pay a larger price for centre embedding leading us tospeculate that resistance to centre embedding may be a natural by-productof heavy reliance on word order (ie the lsquolsquodark sidersquorsquo of a word orderlanguage) Italians are less affected by centre embedding at least withinthe variations that we have attempted here a fact that may result from theapplication of lsquolsquolocalrsquorsquo solutions in a language that relies heavily onagreement morphology However Italians do pay a signicant cost inprocessing time when agreement must compete with a strong (albeit non-deterministic) word order bias Agreement does lsquolsquowinrsquorsquo in such situationsbut only after a measurable delay We must assume that such dilemmasrarely come up in everyday life for Italian speakers due to theconvergence of lexical prosodic and contextual information that usuallyapplies when non-canonical word orders are used But when competitiondoes occur and other favourite sources of information are not around tosettle the ght Italians pay a greater cost than English speakers (whoreally donrsquot care very much about any source of information other thanword order)

Of course more research is needed to achieve a more completeunderstanding of cross-linguistic similarities and differences in the timecourse of language processing However the cross-language differencesthat we have uncovered here regarding preferred interpretations ofcomplex sentences and their associated processing costs have interesting

112 BATES ET AL

implications not only for future psycholinguistic research but fortypological and diachronic linguistics Theorists like Hawkins (1994)and Givon (1995) have speculated about the role of processingconstraints on historical language change and on the class of languagesthat can be processed in real time Although these hypotheses are wellgrounded in linguistic data and appear to have considerable face validity(based on introspection about what is lsquolsquohardrsquorsquo or lsquolsquoeasyrsquorsquo in onersquos nativelanguage) cross-linguistic studies of on-line sentence processing mayprovide a more solid grounding to speculations about the factors thatforce historical language change The results presented here suggest thatthe costs associated with centre embedding and the costs involved inprocessing of non-canonical word order types are both greater in alanguage that relies heavily on word order Conversely the costsassociated with pockets of morphological ambiguity and real or illusivemorphological competition may be greater in a language that permitsextensive word order variation Hence we may nd that differentlanguages have different lsquolsquobreaking pointsrsquorsquo when the communicativeprocess is placed under pressure in the lifetime of an individual andperhaps in the lifetime of the language itself

Manuscript received April 1998Revised manuscript received September 1998

REFERENCESBates E Friederici A amp Wulfeck B (1987) Comprehension in aphasia A crosslinguistic

study Brain and Language 32 19ndash67Bates E MacWhinney B Caselli CM Devescovi A Natale F amp Venza V (1984) A

cross-linguistic study of the development of sentence interpretation strategies ChildDevelopment 55 341ndash354

Bates E McNew S MacWhinney B Devescovi A amp Smith S (1982) Functionalconstraints on sentence processing A cross-linguistic study Cognition 11 245ndash299

Bates E Wulfeck B amp MacWhinney B (1991) Crosslinguistic research in aphasia Anoverview Brain and Language 41 123ndash148

Bazzanella C (1994) Le facce del parlare Florence La Nuova ItaliaBeninca P (1993) Sintassi In AA Sobrero (Ed) Introduzione allrsquoitaliano contemporaneo

Le Strutture (Introduction to contemporary Italian Structures) (pp 41ndash100) Florence LaNuova Italia

Brunswik E (1956) Perception and the representative design of psychology experimentsBerkeley CA University of California Press

Caplan D amp Hildebrandt N (1988) Disorders of syntactic comprehension CambridgeMA MIT Press

Cohen J MacWhinney B Flatt M amp Provost J (1993) PsyScope A new graphicinteractive environment for designing psychology experiments Behavior ResearchMethods Instrument amp Computers 25 257ndash271


Devescovi A (1992) Sentence comprehension in Italian children from 2 to 9 years of agePaper presented at the International Colloquium on Cross-Linguistic Research on theProcessing Acquisition and Disorders of Language Universite Rene Descartes ParisDecember

Devescovi A Bates E DrsquoAmico S Hernandez A Marangolo P Pizzamiglio L ampRazzano C (1997) Grammaticality judgments in patients with and without expressiveagrammatism Aphasiology 11 543ndash579

Devescovi A DrsquoAmico S Smith S Mimica I amp Bates E (1998) The development ofsentence comprehension in Italian and Serbo-Croatian Local versus distributed cues InD Hillert (Ed) Sentence processing A cross-linguistic perspective (pp 345ndash377) SanDiego CA Academic Press

Devescovi P (1986) Ordine delle parole in italiano Aspetti linguistici e psicologici (Wordorder in Italian Linguistic and psychological aspects) Doctoral dissertation Universitadegli studi di Roma lsquolsquoLa Sapienzarsquorsquo

Duranti A amp Ochs E (1979) Left-dislocation in Italian conversation In T Givon(Ed) Syntax and semantics Discourse and syntax (pp 377ndash416) New York AcademicPress

Gibson E (1992) On the adequacy of the competition model (Reviewof The crosslinguisticstudy of sentence processing) Language 68 812ndash830

Givon T (1995) Functionalism and grammar Philadelphia PA J BenjaminsGregory RL (1966) Eye and brain The psychology of seeing London Weidenfeld amp

NicolsonHawkins J (1994) A performance theory of order and constituency New York Cambridge

University PressHernandez A Bates E amp Avila L (1994) On-line sentence interpretation in Spanishndash

English bilinguals What does it mean to be lsquolsquoin betweenrsquorsquo Applied Psycholinguistics 15417ndash446

Kail M (1989) Cue validity cue cost and processing types in French sentencecomprehension In B MacWhinney amp E Bates (Eds) The crosslinguistic study oflanguage processing (pp 77ndash117) New York Cambridge University Press

Kilborn K (1987) Sentence processing in a second language Seeking a performancedenition of uency PhD dissertation University of California San Diego CA

Kilborn K amp Ito T (1989) Sentence processing strategies in adult bilinguals In BMacWhinney amp E Bates (Eds) The crosslinguistic study of sentence processing (pp 257ndash291) New York Cambridge University Press

Kimball J (1973) Seven principles of surface structure parsing in natural languageCognition 2 15ndash47

Li P Bates E amp MacWhinney B (1993) Processing a language without inections Areaction time study of sentence interpretation in Chinese Journal of Memory andLanguage 32 169ndash192

Liu H Bates E amp Li P (1992) Sentence interpretation in bilingual speakers of Englishand Chinese Applied Psycholinguistics 13 451ndash484

MacWhinney B amp Bates E (Eds) (1989) The crosslinguistic study of sentence processingNew York Cambridge University Press

MacWhinney B Bates E amp Kliegl R (1984) Cue validity and sentence interpretation inEnglish German and Italian Journal of Verbal Learning and Verbal Behaviour 23 127ndash150

MacWhinney B amp Pleh C (1988) The processing of restrictive relative clauses inHungarian Cognition 29 95ndash141

Massaro D (1987) Speech perception by ear and eye A paradigm for psychological inquiryHillsdale NJ Lawrence Erlbaum Associates Inc

114 BATES ET AL

Sheldon A (1974) On the role of parallel function in the acquisition of relative clauses inEnglish Journal of Verbal Learning and Verbal Behavior 13 272ndash281

Simone R (1990) Fondamenti di linguistica (Foundations of linguistics) Bari La TerzaSimone R (1993) Stabilitarsquo e instabilitarsquo nei caratteri originali dellrsquoitaliano [Stability and

instability in original characters in Italian] In AA Sobrero (Ed) Introduzione allrsquoitalianocontemporaneo Le Strutture [Introduction to contemporary Italian Structures] (pp 247ndash290) Florence La Nuova Italia

Smith S amp Bates E (1987) Accessibility of case and gender contrasts for agentndashobjectassignment in Brocarsquos aphasics and uent anomics Brain and Language 30 8ndash32

Smith S amp Mimica I (1984) Agrammatism in a case-inected language Comprehensionof agentndashobject relations Brain and Language 21 274ndash290

Wulfeck B Bates E amp Capasso R (1991) A crosslinguistic study of grammaticalityjudgments in Brocarsquos aphasia Brain and Language 41 311ndash336

APPENDIX 1ItalianEnglish relative clause experiment Nouns and verbs used

Italian English Italian English

N1S (animals) N1S (animals) N1P (animals) N1P (animals)il cane the dog i cani the dogsla puzzola the skunk le puzzole the skunksla talpa the mole le talpe the molesil cavallo the horse i cavalli the horsesil canguro the kangaroo i canguri the kangaroosil gatto the cat i gatti the catslrsquoelefante the elephant gli elefanti the elephantsil leone the lion i leoni the lionsla tigre the tiger le tigri the tigersla mucca the cow le mucche the cowsla foca the seal le foche the sealsla giraffa the giraffe le giraffe the giraffes

N2S (humans) N2S (humans) N2P (humans) N2P (humans)il falegname the carpenter i falegnami the carpentersla maestra the teacher le maestre the teachersla segretaria the secretary le segretarie the secretariesil dottore the doctor i dottori the doctorsil parrucchiere the hairdresser i parrucchieri the hairdressersil professore the professor i professori the professorsla cameriera the waitress le cameriere the waitressesil cuoco the cook i cuochi the cooksil contadino the farmer i contadini the farmersla ballerina the ballerina le ballerine the ballerinas


ItalianEnglish relative clause experiment Nouns and verbs used


il soldato the soldier i soldati the soldiersil pagliaccio the clown i pagliacci the clownsil dentista the dentist i dentisti the dentists

V1S V1S V1P V1Ppicchia strikes picchiano strikemorde bites mordono bitetira shoots tirano shootcolpisce hits colpiscono hitrompe breaks rompono breakgrafa scratches grafano scratchuccide murders uccidino murderpizzica pinches pizzicano pinchbrucia burns bruciano burnsuffoca suffocates soffocano suffocatespinge pushes spingono pushstrozza strangles strozzano strangledistrugge destroys distruggono destroyammazza kills ammazzano killbastona beats bastonano beat

V2S V2S V2P V2Pvede sees vedono seeaspetta expects aspettano expectsente hears sentono hearascolta listens to ascoltano listen toguarda watches guardano watchchiama calls chiamano callcerca seeks cercano seeksegue follows seguono followosserva observes osservano observe

APPENDIX 2Sample English Sentences for Experiments 1ndash4a b

Experiments 1 and 3 Word order variations

1a3a Simple sentences only

NVN The secretary shoots the cookNNV The secretary the cook shootsVNN Shoots the secretary the cook

1b3b Main clause interpretation

NVN The secretary that sees the doctor shoots the cook (rel pos 1VN)The secretary that the doctor sees shoots the cook (rel pos 1NV)

116 BATES ET AL

The secretary shoots the cook that sees the doctor (rel pos 2NV)The secretary shoots the cook that the doctor sees (rel pos 1VN)

NNV The secretary that sees the doctor the cook shoots (rel pos 1VN)The secretary that the doctor sees the cook shoots (rel pos 1NV)The secretary the cook that sees the doctor shoots (rel pos 2VN)The secretary the cook that the doctor sees shoots (rel pos 2NV)

VNN Shoots the secretary that sees the doctor the cook (rel pos 1VN)Shoots the secretary that the doctor sees the cook (rel pos 1NV)Shoots the secretary the doctor that sees the cook (rel pos 2VN)Shoots the secretary the doctor that the cook sees (rel pos 1NV)

1c3c Relative clause interpretation

NVN The secretary that shoots the doctor sees the cook (rel pos 1VN)The secretary that the doctor shoots sees the cook (rel pos 1NV)The secretary sees the cook that shoots the doctor (rel pos 2VN)The secretary sees the cook that the doctor shoots (rel pos 2NV)

NNV The secretary that shoots the doctor the cook sees (rel pos 1VN)The secretary that the doctor shoots the cook sees (rel pos 1NV)The secretary the cook that shoots the doctor sees (rel pos 2VN)The secretary the cook that the doctor shoots sees (rel pos 2NV)

VNN Sees the secretary that shoots the doctor the cook (rel pos 1VN)Sees the secretary that the doctor shoots the cook (rel pos 1NV)Sees the secretary the doctor that shoots the cook (rel pos 2VN)Sees the secretary the doctor that the cook shoots (rel pos 2NV)

Experiments 2 amp 4 Agreement Variations3


Ag0 The secretary shoots the doctorAg1 The secretary shoots the doctorsAg3 The secretary shoot the doctors


Ag0 The secretary that sees the doctor shoots the cook (rel pos 1VNAg0rel)main The secretary that the doctor sees shoots the cook (rel pos 1NVAg0rel)

The secretary that sees the doctors shoots the cook (rel pos 1VNAg1rel)The secretary that the doctors sees shoots the cook (rel pos 1NVAg1rel)The secretary that see the doctors shoots the cook (rel pos 1VNAg2rel)The secretary that the doctors see shoots the cook (rel pos 1NVAg2rel)

The secretary shoots the cook that sees the doctor (rel pos 2VNAg0rel)The secretary shoots the cook that the doctor sees (rel pos 2NVAg0rel)The secretary shoots the cook that sees the doctors (rel pos 2VNAg1rel)The secretary shoots the cook that the doctors sees (rel pos 2NVAg1rel)The secretary shoots the cook that see the doctors (rel pos 2VNAg2rel)The secretary shoots the cook that the doctors sees (rel pos 2NVAg2rel)

Ag1 The secretary that sees the doctor shoots the cooks (rel pos 1VNAg0rel)main The secretary that the doctor sees shoots the cook (rel pos 1NVAg0rel)

The secretary that sees the doctors shoots the cooks (rel pos 1VNAg1rel)

PROCESSING COMPLEX SENTENCES 117The secretary that the doctors sees shoots the cooks (rel pos 1NVAg1rel)The secretary that see the doctors shoots the cooks (rel pos 1VNAg2rel)The secretary that the doctors see shoots the cooks (rel pos 1NVAg2rel)

The secretary shoots the cooks that sees the doctor (rel pos 2VNAg0rel)The secretary shoots the cooks that the doctor sees (rel pos 2NVAg0rel)The secretary shoots the cooks that sees the doctors (rel pos 2VNAg1rel)The secretary shoots the cooks that the doctors sees (rel pos 2NVAg1rel)The secretary shoots the cooks that see the doctors (rel pos 2VNAg2rel)The secretary shoots the cooks that the doctors see (rel pos 2NVAg2rel)

Ag2 The secretary that sees the doctor shoot the cooks (rel pos 1VNAg0rel)main The secretary that the doctor sees shoot the cooks (rel pos 1NVAg0rel)

The secretary that sees the doctors shoot the cooks (rel pos 1VNAg1rel)The secretary that the doctors sees shoot the cooks (rel pos 1NVAg1rel)The secretary that see the doctors shoot the cooks (rel pos 1VNAg2rel)The secretary that the doctors see shoot the cooks (rel pos 1NVAg2rel)

The secretary shoot the cooks that sees the doctor (rel pos 2VNAg0rel)The secretary shoot the cooks that the doctor sees (rel pos 2NVAg0rel)The secretary shoot the cooks that sees the doctors (rel pos 2VNAg1rel)The secretary shoot the cooks that the doctors sees (rel pos 2NVAg1rel)The secretary shoot the cooks that see the doctors (rel pos 2VNAg2rel)The secretary shoot the cooks that the doctors see (rel pos 2NVAg2rel)


Ag0 The secretary that shoots the doctor sees the cook (rel pos 1VNAg0rel)main The secretary that the doctor shoots sees the cook (rel pos 1NVAg0rel)

The secretary that shoots the doctors sees the cook (rel pos 1VNAg1rel)The secretary that the doctors shoots sees the cook (rel pos 1NVAg1rel)The secretary that shoot the doctors sees the cook (rel pos 1VNAg2rel)The secretary that the doctors shoot sees the cook (rel pos 1NVAg2rel)

The secretary sees the cook that shoots the doctor (rel pos 2VNAg0rel)The secretary sees the cook that the doctor shoots (rel pos 2NVAg0rel)The secretary sees the cook that shoots the doctors (rel pos 2VNAg1rel)The secretary sees the cook that the doctors shoots (rel pos 2NVAg1rel)The secretary sees the cook that shoot the doctors (rel pos 2VNAg2rel)The secretary sees the cook that the doctors shoot (rel pos 2NVAg2rel)

Ag1 The secretary that shoots the doctor sees the cooks (rel pos 1VNAg0rel)main The secretary that the doctor shoots sees the cooks (rel pos 1NVAg0rel)

The secretary that shoots the doctors sees the cooks (rel pos 1VNAg1rel)The secretary that the doctors shoots sees the cooks (rel pos 1NVAg1rel)The secretary that shoot the doctors sees the cooks (rel pos 1VNAg2rel)The secretary that the doctors shoot sees the cooks (rel pos 1NVAg2rel)

The secretary sees the cooks that shoots the doctor (rel pos 2VNAg0rel)The secretary sees the cooks that the doctor shoots (rel pos 2NVAg0rel)The secretary sees the cooks that shoots the doctors (rel pos 2VNAg1rel)The secretary sees the cooks that the doctors shoots (rel pos 2NVAg1rel)The secretary sees the cooks that shoot the doctors (rel pos 2VNAg2rel)The secretary sees the cooks that the doctors shoots (rel pos 2NVAg2rel)

118 BATES ET AL

Ag2 The secretary that shoots the doctor see the cooks (rel pos 1VNAg0rel)main The secretary that the doctor shoots see the cooks (rel pos 1NVAg0rel)

The secretary that shoots the doctors see the cooks (rel pos 1VNAg1rel)The secretary that the doctors shoots see the cooks (rel pos 1NVAg1rel)The secretary that shoot the doctors see the cooks (rel pos 1VNAg2rel)The secretary that the doctors shoot see the cooks (rel pos 1NVAg2rel)

The secretary see the cooks that shoots the doctor (rel pos 2VNAg0rel)The secretary see the cooks that the doctor shoots (rel pos 2NVAg0rel)The secretary see the cooks that shoots the doctors (rel pos 2VNAg1rel)The secretary see the cooks that the doctors shoots (rel pos 2NVAg1rel)The secretary see the cooks that shoot the doctors (rel pos 2VNAg2rel)The secretary see the cooks that the doctors shoot (rel pos 2NVAg2rel)

aFor the sake of economy only English examples are listed Italian items are directtranslations with verbs in the present indicative (eg lsquolsquoshootsrsquorsquo = spara lsquolsquoshootrsquorsquo = sparano)

bThe same lexical items are used in all examples to facilitate comparison actual stimulicontain ve different randomised assignments of lexical items to conditions (see Table 1)

cThese examples are restricted to cases in which the rst noun is singular and the secondnoun is plural on all items with an agreement contrast in the full set of stimuli plural vssingular nouns and verbs are counterbalanced across all agreement conditions

Abbreviation rel pos = relative position

APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123

70 BATES ET AL

(MacWhinney amp Bates 1989) The term lsquolsquocompetitionrsquorsquo reects a centralassumption of the model That different sources of information (ielsquolsquocuesrsquorsquo) converge compete andor conspire to determine the outcome ofsentence processing with different outcomes depending on the relativestrength of cues from one language to another For example the model isdesigned to capture the fact that two languages with the same basic wordorder (eg subjectndashverbndashobject or SVO) can differ markedly in the extentto which listeners rely on word order information to assign semantic rolesin sentence comprehension compared with other sources of informationlike subjectndashverb agreement

In a long series of sentence comprehension studies researchers workingwithin the Competition Model have shown that listeners rely on the mostvalid (ie frequent and reliable) sources of information in their languageMacWhinney and Bates (1989) refer to this effect as lsquolsquocue validityrsquorsquo a termborrowed from Brunswik (1956) English is a language in which word orderis high in cue validity Constituent order is rigidly preserved acrosssentence types and correlates highly with semantic roles (ie lsquolsquowho didwhat to whomrsquorsquo) Italian is in the same typological category (SVO withoutcase marking on nouns) but word order is low in cue validity that isbecause extensive variation of word order is permitted for pragmaticpurposes the correlation between word order and semantic roles isrelatively low Studies of sentence comprehension in these two languageshave shown that English listeners rely primarily on word order to decidelsquolsquowho did the actionrsquorsquo making little use of subjectndashverb agreement orsemantic contrasts Italian listeners are heavily inuenced by subjectndashverbagreement and semantic contrasts but they pay little attention to wordorder The same paradigm has now been used in more than a dozenlanguages (English Italian Spanish French German Dutch Serbo-Croatian Hungarian Japanese Chinese Warlpiri Hebrew ArabicGreek for summaries see Li Bates amp MacWhinney 1993 MacWhinneyamp Bates 1989) It is clear from these studies that many different proles orhierarchies of cue utilization are possible (eg word order gt agreement gtsemantics in English case gt agreement gt semantics gt word order in Serbo-Croatian passive marking gt semantics gt word order gt topic marking inChinese) For any given language the most valid cues also tend to be therst ones used by children (Devescovi et al 1998 Kail 1989) the mostprone to transfer during second-language learning (Hernandez Bates ampAvila 1994 Kilborn amp Ito 1989 Liu Bates amp Li 1992) and the mostresistant to loss following focal brain injury (Bates Wulfeck ampMacWhinney 1991)

In most of these studies listeners are asked to lsquolsquochoose the one who didthe actionrsquorsquo in response to simple sentences or sentence-like stringsconsisting of a transitive verb and two concrete animate or inanimate













72 BATES ET AL







74 BATES ET AL









76 BATES ET AL








78 BATES ET AL












80 BATES ET AL











82 BATES ET AL









84 BATES ET AL










86 BATES ET AL










(NVN(VN))



88 BATES ET AL







AmbiguousVN 880 880 820NV 800 735 925







90 BATES ET AL











92 BATES ET AL










94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123













72 BATES ET AL







74 BATES ET AL









76 BATES ET AL








78 BATES ET AL












80 BATES ET AL











82 BATES ET AL









84 BATES ET AL










86 BATES ET AL










(NVN(VN))



88 BATES ET AL







AmbiguousVN 880 880 820NV 800 735 925







90 BATES ET AL











92 BATES ET AL










94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123





74 BATES ET AL









76 BATES ET AL








78 BATES ET AL












80 BATES ET AL











82 BATES ET AL









84 BATES ET AL










86 BATES ET AL










(NVN(VN))



88 BATES ET AL







AmbiguousVN 880 880 820NV 800 735 925







90 BATES ET AL











92 BATES ET AL










94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123







76 BATES ET AL








78 BATES ET AL












80 BATES ET AL











82 BATES ET AL









84 BATES ET AL










86 BATES ET AL










(NVN(VN))



88 BATES ET AL







AmbiguousVN 880 880 820NV 800 735 925







90 BATES ET AL











92 BATES ET AL










94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123





78 BATES ET AL












80 BATES ET AL











82 BATES ET AL









84 BATES ET AL










86 BATES ET AL










(NVN(VN))



88 BATES ET AL







AmbiguousVN 880 880 820NV 800 735 925







90 BATES ET AL











92 BATES ET AL










94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123






80 BATES ET AL











82 BATES ET AL









84 BATES ET AL










86 BATES ET AL










(NVN(VN))



88 BATES ET AL







AmbiguousVN 880 880 820NV 800 735 925







90 BATES ET AL











92 BATES ET AL










94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123






82 BATES ET AL









84 BATES ET AL










86 BATES ET AL










(NVN(VN))



88 BATES ET AL







AmbiguousVN 880 880 820NV 800 735 925







90 BATES ET AL











92 BATES ET AL










94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123





84 BATES ET AL










86 BATES ET AL










(NVN(VN))



88 BATES ET AL







AmbiguousVN 880 880 820NV 800 735 925







90 BATES ET AL











92 BATES ET AL










94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123





86 BATES ET AL










(NVN(VN))



88 BATES ET AL







AmbiguousVN 880 880 820NV 800 735 925







90 BATES ET AL











92 BATES ET AL










94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123






(NVN(VN))



88 BATES ET AL







AmbiguousVN 880 880 820NV 800 735 925







90 BATES ET AL











92 BATES ET AL










94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123

88 BATES ET AL







AmbiguousVN 880 880 820NV 800 735 925







90 BATES ET AL











92 BATES ET AL










94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123





90 BATES ET AL











92 BATES ET AL










94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123






92 BATES ET AL










94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123






94 BATES ET AL









96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123






96 BATES ET AL









98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123






98 BATES ET AL








100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123




100 BATES ET AL














102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123









102 BATES ET AL










104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123







104 BATES ET AL













(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123






(b)

(a)

106







108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123






108 BATES ET AL








110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123





110 BATES ET AL









112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123






112 BATES ET AL


































114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123






















114 BATES ET AL























116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123













116 BATES ET AL





























118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123













118 BATES ET AL








APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123

APP

END

IX3

Off

-lin

ech

oice

AN

OV

AS

igni

fican

tres

ults

a

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

1as

impl

ese

nten

ces

lang

uage

148

161

40

0001

wor

dor

der

296

717

30

0001

248

713

60

0001

248

138

10

0001

lang

uage

wor

dor

der

296

112

10

0001

1bm

ain

clau

sela

ngua

ge1

4826

85

000

01m

ain

clau

seor

der

(MC

O)

296

609

90

0001

248

660

30

0001

248

816

000

1re

lativ

ecl

ause

posi

tion

(RC

P)1

489

680

003

124

159

60

001

lang

uage

MC

O2

9614

58

000

01M

CO

RC

P2

968

260

001

248

202

80

0001

lang

uage

MC

OR

CP

296

713

000

1

1cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

664

000

22

485

320

008

rela

tive

clau

seor

der

(RC

O)

148

987

20

0001

124

610

50

0001

124

816

000

01la

ngua

geR

CO

148

874

000

5M

CO

RC

O2

969

90

003

248

103

70

0001

lang

uage

MC

OR

CO

296

631

000

3

2as

impl

ese

nten

ces

lang

uage

148

441

60

0001

agre

emen

t2

9641

26

000

012

4842

98

000

01la

ngua

geag

reem

ent

296

253

60

0001

119

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123

APP

END

IX3

(con

t)O

ff-l

ine

choi

ceA

NO

VA

Sig

nific

antr

esul

tsa

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

2bm

ain

clau

sela

ngua

ge1

4847

48

000

01m

ain

clau

seag

reem

ent

(MC

A)

296

526

20

0001

248

534

30

0001

rela

tive

clau

seor

der

(RC

O)

124

118

10

002

lang

uage

MC

A2

9630

39

000

01la

ngua

geR

CO

148

831

000

6R

CO

rela

tive

clau

sepo

sitio

n(R

CP)

124

125

10

002

rela

tive

clau

seag

reem

ent

(RC

A)

RC

O2

965

920

004

MC

AR

CA

RC

O4

192

358

000

44

964

162

000

4la

ngua

geM

CA

RC

AR

CO

419

23

963

000

4

2cr

elat

ive

clau

sela

ngua

ge1

487

230

01m

ain

clau

seag

reem

ent

(MC

A)

296

728

000

12

486

630

003

rela

tive

clau

seag

reem

ent

(RC

A)

296

100

550

0001

248

135

270

0001

rela

tive

clau

seor

der

(RC

O)

196

535

130

0001

124

498

770

0001

124

822

40

0001

rela

tive

clau

sepo

sitio

n(R

CP)

124

905

000

6la

ngua

geR

CA

296

650

60

0001

lang

uage

RC

O1

9614

248

000

01la

ngua

geR

CP

148

104

30

002

MC

AR

CA

419

24

120

003

496

578

000

01R

CA

RC

O2

9628

23

000

012

4828

44

000

01R

CA

RC

P2

965

940

004

248

585

000

5la

ngua

geM

CA

RC

O4

192

559

000

01la

ngua

geR

CA

RC

O2

9620

29

000

01aE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

120

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123

APP

END

IX4

Off

-lin

eve

rsus

on-li

nech

oice

AN

OV

AS

igni

fican

tres

ults

invo

lvin

gth

efa

ctor

lsquolsquoexp

erim

entrsquorsquo

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

expe

rim

ent

196

720

001

148

823

001

expe

rim

ent

wor

dor

der

296

112

20

0001

lang

uage

expe

rim

ent

wor

dor

der

219

24

720

01

3bm

ain

clau

seex

peri

men

tm

ain

clau

seor

der

219

26

870

001

expe

rim

ent

mai

ncl

ause

orde

rpo

sitio

n2

966

450

002

3cr

elat

ive

clau

seex

peri

men

t1

9612

41

000

11

4810

96

000

2la

ngua

geex

peri

men

t1

968

530

004

expe

rim

ent

rela

tive

clau

seor

der

148

749

000

9

4as

impl

ese

nten

ces

4bm

ain

clau

se

4cr

elat

ive

clau

seex

peri

men

tre

lativ

ecl

ause

posi

tion

146

147

50

001

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

clau

sepo

sitio

n1

4610

14

000

3

expe

rim

ent

rela

tive

clau

seag

reem

ent

rela

tive

292

619

20

003

clau

seor

der

rela

tive

clau

sepo

sitio

naE

xper

imen

t-w

ide

alph

ale

vels

etat

Plt

001

121

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123

APP

END

IX5

On-

line

reac

tion

time

AN

OV

AS

igni

fica

ntre

sults

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

3as

impl

ese

nten

ces

wor

dor

der

296

384

70

0001

lang

uage

wor

dor

der

296

770

001

3bm

ain

clau

sem

ain

clau

seor

der

(MC

O)

296

822

60

0001

248

609

10

0001

248

238

70

0001

rela

tive

clau

sepo

sitio

n(R

CP)

148

743

000

91

2415

80

000

1re

lativ

ecl

ause

orde

r(R

CO

)1

4830

98

000

011

2440

19

000

01la

ngua

geM

CO

296

673

000

2la

ngua

geR

CO

148

911

000

4la

ngua

geM

CO

RC

O2

964

970

009

MC

OR

CP

296

826

000

12

4890

46

000

12

4822

26

000

1la

ngua

geM

CO

RC

P2

9611

25

000

01

3cr

elat

ive

clau

sem

ain

clau

seor

der

(MC

O)

296

485

00

0001

248

430

70

0001

248

142

70

0001

rela

tive

clau

seor

der

(RC

O)

148

303

30

0001

124

177

00

0001

124

156

90

001

MC

OR

CO

296

748

000

12

488

410

001

rela

tive

clau

sepo

sitio

n(R

CP)

148

146

80

0001

248

181

70

0001

MC

OR

CP

296

426

50

0001

248

486

60

0001

248

770

000

1la

ngua

geM

CO

RC

P2

966

240

003

RC

OR

CP

148

875

000

51

2410

18

000

4M

CO

RC

OR

CP

296

117

50

0001

248

549

000

72

486

510

003

4as

impl

ese

nten

ces

lang

uage

144

188

70

0001

122

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123

APP

END

IX5

(con

t)O

n-lin

ere

actio

ntim

eA

NO

VA

Sig

nific

antr

esul

ts

Ove

rall

Eng

lish

only

Ital

ian

only

Exp

erim

ent

Sour

cedf

FP

dfF

Pdf

FP

4bm

ain

clau

sela

ngua

ge1

4414

65

000

01m

ain

clau

seag

reem

ent

(MC

A)

288

585

80

0001

244

768

000

12

4452

56

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4422

43

000

011

2216

20

000

01la

ngua

geM

CA

288

249

90

001

MC

Are

lativ

ecl

ause

agre

emen

t(R

CA

)4

164

500

000

14

884

550

002

RC

AR

CO

288

129

20

0001

244

97

000

012

446

640

003

rela

tive

clau

sepo

sitio

n(R

CP)

144

200

650

0001

122

224

210

0001

122

376

20

0001

lang

uage

RC

P1

4421

22

000

01R

CA

RC

OR

CP

244

157

10

0001

lang

uage

RC

AR

CP

288

630

50

003

lang

uage

RC

AR

CO

RC

P2

887

152

000

1

4cr

elat

ive

clau

sela

ngua

ge1

449

930

003

mai

ncl

ause

agre

emen

t(M

CA

)2

886

890

002

244

936

000

01re

lativ

ecl

ause

agre

emen

t(R

CA

)2

889

300

002

244

954

000

01re

lativ

ecl

ause

orde

r(R

CO

)1

4435

90

000

011

2216

27

000

11

2222

76

000

01re

lativ

ecl

ause

posi

tion

(RC

P)1

4436

492

000

011

2228

14

000

011

2211

81

000

02la

ngua

geR

CA

288

654

000

2R

CA

RC

O2

8839

27

000

012

4412

57

000

012

4428

19

000

01la

ngua

geR

CA

RC

O2

889

230

0001

RC

OR

CP

144

242

20

0001

122

334

50

0001

RC

OR

CA

RC

P2

888

780

0001

lang

uage

RC

OR

CP

144

242

20

0001

lang

uage

RC

OR

CA

RC

P2

885

340

006

Exp

erim

ent-

wid

eal

pha

leve

lset

atP

lt0

01

123

processing complex sentences: a cross-linguistic study · pdf fileprocessing complex...

Documents