Relative clause extractions generally yield so-called island effects (degraded judgments) across languages (Ross, 1967), (1). However, Swedish and the other Mainland Scandinavian languages comprise famous exceptions to this pattern (Engdahl and Ejerhed, 1982), cf (2):

(1) * Those books I know a girl that writes.

(2) S ̊ana bo ̈cker ka ̈nner jag en tjej som skriver. such books know I a girl who writes

The origin of island effects is under debate. Sprouse et al. (2012) found superadditive effects in judgments of island violations in English – combined effects greater than the sum of the individual costs for extraction and complexity (island structure). Because no correlation was found between participants’ working memory (WM) span (measured via serial recall and n-back) and the superadditive effects, they concluded that island effects must be deri- ved from violations of syntactic constraints rather than processing limitations. In contrast, Hofmeister et al. (2014) did find a correlation between WM span (measured via reading span) and superadditivity in a rapid serial visual presentation experiment. They also found superadditive effects in fully grammatical, but hard to process sentences, suggesting that processing factors do play a role in superadditivity and hence in island effects.

In our talk, we report on the results from an acceptability judgment experiment that was designed to look for superadditive effects in relative clause extractions in Swedish. We investigated the (super)additive effect of clause type (that-clause (3a), (3c), vs. relative clause, (3b), (3d)), and extraction type (short, (3a)–(3b), vs long, (3c)–(3d)). We also tested whether any such effects correlate with WM span, as measured via an n-back and a reading span task. As a control condition, we also included garden-path sentences and their non- garden-path counterparts, as the processing of garden-paths has been shown to correlate with performance on n-back (Novick et al., 2014).(3)

a. Det var Anna som anma ̈lde att en kille snattade s ̊ana chokladkakor it was Anna who reported that a guy stole such chocolate bars i godisaffa ̈ren.from candy store-the

 b.Det var Anna som anma ̈lde en kille som snattade s ̊ana chokladkakor iit was Anna who reported a guy who stole such chocolate bars from godisaffa ̈ren.candy store-the

c. Det var s ̊ana chokladkakor som Anna anma ̈lde att en kille snattade i it was such chocolate bars that Anna reported that a guy stole in godisaffa ̈ren.candy store-the

d.Det var s ̊ana chokladkakor som Anna anma ̈lde en kille som snattade i it was such chocolate bars that Anna reported a guy who stole in godisaffa ̈ren.candy store-the

Unexpectedly, we found superadditive effects for the relative clause extractions, but no cor- relation with WM, and significantly different ratings for garden-path and non-garden-path sentences, but again no relation with WM. However, since there was no relation between the ratings of garden-paths and WM it is evident that the relation between offline acceptability judgements and WM measure is more complex than assumed in the literature. We therefore cannot conclude that the lack of correlation between the superadditive effect and WM for the extraction data is due to syntactic violations. Rather, we conclude that what causes superadditive effects is still an open question that requires further investigation.

Referenser

Engdahl, Elisabet, and Eva Ejerhed, ed. 1982. Readings on unbounded dependencies in Scandinavian languages. Ume ̊a: Ume ̊a University.

Hofmeister, Philip, Laura Staum Casasanto, and Ivan A. Sag. 2014. Processing effects in lin- guistic judgment data: (super-)additivity and reading span scores. Language and Cogni- tion 6:111–145. URL http://journals.cambridge.org/articleS1866980813000070.

Novick, Jared M., Erika Hussey, Susan Teubner-Rhodes, J. Isaiah Harbison, and Michael F. Bunting. 2014. Clearing the garden-path: improving sentence processing through cogni- tive control training. Language, Cognition and Neuroscience 29:186–217.

Ross, John Robert. 1967. Constraints on variables in syntax. Doctoral Dissertation, MIT, Cambridge, Mass. [published as ‘Infinite syntax’. 1986].

Sprouse, Jon, Matt Wagers, and Colin Phillips. 2012. A test of the relation between working- memory capacity and syntactic island effects. Language 88:82–123.

The focus in this study is methods for screening young children’s, subitizing abilities. Subitize is in this project defined as identifying the number of things in a setsimply by quickly looking at them and without counting them one by one. The aim of the study is to compare two different screening instruments for identifyingchildren’s subitizing abilities, individually and in groups of children, in a Swedish preschool class context. In total, data from 44 children is collected, who havetaken one group test (Ability to Quickly See Quantity - AQSQ), and an individual test (Number Set Test - NST). The research question was if the children’sresults correlate with each other, or not, no matter of test conditions (individual/group). The results show a moderate to strong positive correlation, usingPearson and T-test Pair Samples Correlation (r=.559 p=

Swedish relative clause (RC) extractions are allegedly acceptable despite a universal constraint on NP extraction (Erteschik-Shir, 1973). In two eye-tracking experiments, we investigated the processing of RC extractions in Swedish by manipulating the plausibility of NP fillers relative to RC predicates. A similar study on English RCs (Traxler & Pickering, 1996) found plausibility effects in English for non-island but not island conditions, indicating a lack of integration in islands.

Experiment 1 tested for plausibility effects in extractions from RCs (alleged non-islands), that-clauses (non-islands), and non-restrictive RCs (islands; Engdahl, 1997). We expected that only the first two would show sensitivity to plausibility. Surprisingly, we found effects, and thus integration for all three structures. Experiment 2 tested for plausibility effects in RCs embedded in subjects (islands) vs. objects (non-islands). The RC verbs (optionally transitive) provided a potential gap and were hypothesized to allow for integration in the object condition only. A following clause provided the real gap. Again, we found plausibility effects for both conditions at the RC verb, suggesting integration. Our study calls into question the correspondence between acceptability- and processing-based metrics of Swedish islandhood and raises the possibility of cross-linguistic variation with regard to the processing of island violations.

It has been argued that the Mainland Scandinavian languages differ from English in that they permit filler-gap dependencies into adjunct clauses (Anward, 1982; Hagström, 1976), which are standardly treated as strong islands for extraction(Huang, 1982). A closer look suggests that extraction from adjunct clauses is also possible in English, provided that a coherence relation (e.g., a causal, as opposed to a purely temporal relation) holds between the events referred to by the matrix and the adjunct clause (Truswell, 2011). The same restriction appears to hold in Swedish. However, English contrasts with Swedish in that it has been argued to disallow extraction from finite adjuncts. Observations like these raise important questions regarding filler-gap association in island domains and cross-linguistic variation. First, if coherence matters for extraction from adjunct clauses, then adjuncts appear to contrast with other island domains, in which filler-gap association has been claimed to be suspended, such as in relative clause islands in English (e.g.,Traxler & Pickering, 1996). Second, if finiteness is a potential locus of variation, then this must be explained and related to other evidence of cross-linguistic and inter-individual variation (Kush et al., 2017; Phillips, 2013; Sprouse et al., 2016). 

We conducted two acceptability judgment experiments, one for English and one for Swedish, to investigate the hypothesis that semantic coherence facilitates extraction in both languages, whereas finiteness degrades extraction, but only in English. In our materials, we first manipulated the telicity of the matrix verb such that sentential coherence was either augmented(coherent, 1a and 2a) or impeded(non-coherent, 1band 2b), and then manipulated the finiteness of the adjunct clause ([finite/non-finite] in 1and 2). Long-distance dependencies were tested in the form of question formation in English (1) and topicalization in Swedish (2).

Forty sets of items were created and distributed across four presentation lists, interspersed with 80distractor sentences. Fifty-seven Swedish and 72 English mono-lingual, native speakers rated the sentences using a 7-point Likert scale(1 = “completely unacceptable” to 7 = “completely acceptable”), as presented using Google forms.The z-score adjusted responses for each condition in the two languages are presented in Figure1 and 2. Linear mixed models were used to analyze normalized z-score responses. As expected, coherence significantly improved extraction for both Swedish (β= 0.191, t = 4.792, p<.001) and English (β= 0.121, t = 3.853, p< .001). Furthermore, a significant finiteness by coherence interaction was found for English, in which finiteness was seen to degrade coherent, but not non-coherent structures(β= -0.050, t = -2.159, p< .05). Swedish showed no effect of finiteness or finiteness by coherence interaction.

Our results demonstrate that coherence improves extraction from adjuncts in both Swedish and English, which suggests that filler-gap association is not suspended in adjunct clauses. Furthermore, we show that finiteness reduces the acceptability of coherent structures, but only in English, thus pointing to one case of cross-linguistic variation.One possible explanation for this variation is that finite and non-finite adjuncts in English differ in terms of syntactic complexity: Non-finite after-adjuncts in English have a participial structure and can be assumed to lack a contentful CP-domain. Extraction from such structures thus requires fewer intermediate steps in successive-cyclic movement and a corresponding lower processing cost than extraction from their finite counterparts(see Wurmbrand, to appear, for a similar account for quantifier raising in English). In contrast, Swedish non-finite after-adjuncts involve a complementizer and an infinitive and can be assumed to be similar in structural size to their finite counterparts, requiring equally many movement steps and hence similar processing costs for extraction.

1) 

a. coherent | non-finite/finite

Which beer did he almost stumble [after chugging / after he chugged]?

b. non-coherent | non-finite/finite

Which beer did he stroll a little [after chugging / after he chugged]?

(2) 

a. coherent | non-finite/finite

Den medicinen däckade han alltid [efter att ha tagit / efter att han hade tagit].

that medicine fell-asleep he always after to have taken / after that he had taken

‘He always fell asleep {after taking/after he took} that medicine.’

b. non-coherent | non-finite/finite

Den medicinen stannade han lite [efter att ha tagit / efter att han hade tagit].

that medicine stayed he a little after to have taken / after that he had taken

‘He stayed a little {after taking/after he took} that medicine.’

Figure 1. z-adjusted ratings for Exp. 1 (English) 

Figure 2. z-adjusted ratings for Exp. 2 (Swedish)

References

Anward, J. (1982). Basic Swedish. In E. Engdahl & E. Ejerhed (Eds.), Readings on unbounded dependencies in Scandinavian languages (pp. 47-75). Stockholm: Almqvist & Wiksell.

Hagström, B. (1976). Om satsflätans förutsättningar. En skiss. In L. Svensson, A. M. Wieselgren, & Å. Hansson (Eds.), Nordiska studier i filologi och lingvistik. Festskrift tillägnad Gösta Holm på 60-årsdagen den 8 juli 1976 (pp. 138–150). Lund: Studentlitteratur.

Huang, J. C.-T. (1982). Logical Relations in Chinese and the Theory of Grammar. PhD. Dissertation, MIT, Cambridge, Mass.

Kush, D., Lohndal, T., & Sprouse, J. (2017). Investigating Variation in Island Effects: A Case Study of Norwegian Wh-Extraction. Natural Language & Linguistic Theory. doi:https://doi.org/10.1007/s11049-017-9390-z

Phillips, C. (2013). On the Nature of Island Constraints. II: Language learning and innateness. In J. Sprouse & N. Hornstein (Eds.), Experimental syntax and island effects (pp. 132-157).Cambridge: Cambridge University Press.

Sprouse, J., Caponigro, I., Greco, C., & Cecchetto, C. (2016). Experimental syntax and the variation of island effects in English and Italian. Natural Language & Linguistic Theory, 34, 307–344.

Traxler, M. J., & Pickering, M. J. (1996). Plausibility and the processing of unbounded dependencies: An eye-tracking study. Journal of Memory and Language, 35(3), 454-475.

Truswell, R. (2011). Events, phrases and questions. Oxford: Oxford University Press. 

Wurmbrand, S. (To appear). The cost of raising quantifiers. To appear in Glossa.

This dissertation examines the influence of lexical-semantic representations, conceptual similarity, and contextual fit on the processing of coordinated verb phrases. The study integrates information gleaned from current linguistic theory with current psycholinguistic approaches to examining the processing of coordinated verb phrases. It has been claimed that in coordinated phrases, one conjunct may influence the processing of a second conjunct if they are sufficiently similar. For example, The likelihood of adopting an intransitive analysis for the optionally transitive verb of a subordinated clause in sentences like Although the pirate ship sank the nearby British vessel did not send out lifeboats may be increased if the ambiguous verb (sank) is coordinated with a preceding, intransitively biased verb (halted and sank). Similarly, processing of the second conjunct may be facilitated when coordinated with a similar first conjunct. Such effects, and others in this vein have often been designated “parallelism effects.” However, notions of similarity underlying such effects have long been ill-defined. Many existing studies rely on relatively shallow features like syntactic category information or argument structure generalizations, such as transitive or intransitive, as a basis for structural comparison. But it may be that deeper levels of lexical-semantic representation and more varied, semantic or conceptual sources of information are also relevant to establishing similarity between conjuncts. In addition, little has been done to xi integrate parallelism effects to theories of the processing architecture underlying such effects, particularly for studies involving syntactic ambiguity resolution. Using two word-by-word reading and three eyetracking while reading experiments, I investigate what contribution detailed lexical-semantic representations, as well as conceptual and contextual information make towards establishing parallel coordination in the online processing of coordinated verb phrases. The five studies demonstrate that parallelism effects are indeed sensitive to deeper representational information, conceptual similarity, and contextual fit. Furthermore, by controlling for deeper representational information, it is demonstrated that expected facilitatory patterns arising from coordination of similar conjuncts may be disrupted. Implications for the architecture of the processing system are discussed, and it is argued that constraintbased/ competition models of processing best accommodate the pattern of results.

Within both psycholinguistic theories of parsing and formal theories of syntax, a<br/><br> distinction between arguments and adjuncts is central to some theories, while<br/><br> minimized or denied by others. Even for theories that deem the argument/<br/><br> adjunct distinction important, the exact nature of the distinction has been difficult<br/><br> to characterize. In this article, we review the psycholinguistic evidence for an<br/><br> argument/adjunct distinction, discuss how argument status can best be defined in<br/><br> the light of such evidence, and consider the implications for how grammatical<br/><br> knowledge is represented and accessed in the human mind.

Previous eye tracking research using sentences like (1) suggested that a dative verb introduces an indirect object (recipient) into the discourse (Boland, 2005). Boland found anticipatory fixations on photographs depicting potential recipients following dative (but not transitive) verbs. Datives also require a direct object (theme), but there were relatively few fixations on photographs depicting themes, even after the pronoun. In the current study, the theme attracts a high proportion of fixations when it is prominent (i.e., focused, Selkirk, 1984) in a sentence’s information structure.

In fifteen sets of items we manipulated sentence structure such that, at the verb, the theme was either still in focus (2) or not (3), using the same dative (e.g., loan) and transitive control (e.g., own) verbs in both conditions. We also manipulated the frequency with which the dative verbs occurred with recipient arguments. Twenty-one participants listened to the sentences while looking at photos corresponding to the agent, theme, and potential recipient of the critical verb. A fourth photo depicted an object that was not a potential argument of either the dative or the control verb.

We examined gaze in 100 ms bins 200-1000 ms after the onset of dative and transitive verbs. Fixations on the theme were much more likely in the focused condition. This effect interacted with both verb-type and temporal bin, with the verb-type by focus interaction localized 600-1000 ms after verb onset. For the high frequency datives, fixations on the theme in the focused condition declined 600-1000 ms after verb onset as attention shifted toward other depicted entities.

Fixations on the recipient were also influenced by the focus manipulation, though less dramatically. When the theme was focused, the critical verb was in a relative clause. When the theme was unfocused, the verb was in a more prominent position, as in Boland (2005). The current research investigated whether the anticipatory fixations on recipients found in Boland (2005) would be replicated if the verb were in a less prominent position. In fact, sentence structure interacted with both temporal bin and verb-type. High frequency datives prompted more fixations on recipients across the entire critical interval when the verb was in a main clause. Surprisingly, the control condition also had more fixations on recipients when the verb was in a main clause, but this effect interacted with temporal bin, beginning substantially later than for datives.

This research demonstrates that sentential focus strongly influences visual attention toward event participants, at times minimizing effects of argument structure. In a related experiment, using nearly identical materials in an auditory change- detection paradigm, listeners were less sensitive to dative/transitive verb substitutions when the theme was focused. Taken together, these findings suggest that verb meaning (or perhaps argument structure) is processed more deeply when a verb is in a more prominent position.

Examples

(1) The house was dirty inside, but the realtor swept/sold it easily for/to some newlyweds.

(2) The necklace that the well-known jeweler loaned/owned/sold was worth a million dollars.

(3) Because one comic book was extremely rare, the famous collector loaned/owned/sold it in secret.

References

Boland, J. (2005). Visual arguments. Cognition, 95, 237-274.Selkirk, E. (1984). Phonology and syntax: The relation between sound and structure. Cambridge, MA: MIT Press.

Previous research suggests that “implicit” agent arguments are activated into the discourse during online comprehension of short passive constructions (Mauner & Koenig, 1999). Yet some researchers (e.g., Kratzer, 1996) have claimed that there exists a distinction between stative/adjectival and eventive/verbal passives, and that only the latter is able to license and project an agent. In addition, the two forms are superficially similar in English, which poses a potential problem for L2 learners of English when their L1s use distinct forms to mark whether a passive construction is eventive or stative. One such case is Swedish, where vara (to be) marks the stative and other forms mark the eventive usage.

Using a word-by-word reading paradigm, we tested monolingual English speakers and advanced Swedish L2 learners of English for differences in agent activation between short adjectival (1) and verbal (2) passive constructions. We hypothesized that for native English speakers, verbal passive participles would activate and make available an agent in the discourse, thus facilitating processing at an agent-dependent, rationale clause (to keep...), whereas adjectival passives would not. For the Swedish L2 learners of English, we explored a possible L1/L2 transfer effect (e.g., Frenck-Mestre & Pynte, 1997) whereby superficial similarities between Swedish vara and English to be would support adjectival activation and preclude agent activation, causing processing difficulty for both conditions (1) and (2).

Two processing constraints were manipulated: 1) structural-semantic bias, via a verb or adjective-biasing adverb (very/immediately), and 2) the frequency (BNC) by which a participle appears in its verbal/adjectival form. A mixed model analysis revealed a main effect of structural-semantic bias for which the direction of the effect differed for native speakers and L2 learners. Native speakers had marginally longer RTs at the first verb region of the rational clause and significantly longer RTs at the verb+ region (any spies) for the adjectival passive vs. the verbal passive and a by-phrase control (3). For the L2 learners, both the adjectival and verbal passives had longer RTs than the control, beginning at the verb and through the final regions, suggesting that both forms were processed as adjectivals. No effect of frequency was found for either group.

In sum, these data support the claim that there exists a structural-semantic distinction between verbal and adjectival passives and that for native speakers, only the former can introduce an implicit agent argument to the discourse. In addition, the L2 data suggest that the adjectival form in the L1 may constrain selection to an adjectival form in the L2, which is consistent with the claim that lexical-semantic properties of an L1 may transfer and affect L2 processing, although an additional study is needed to rule out general L2 effects.

Examples

(1) The important document was very classified to keep any spies from discovering the plan.

(2) The important document was immediately classified to keep any spies...

(3) The important document was immediately classified by the general to keep any spies...

References

Frenck-Mestre, C., & Pynte, J. (1997). Syntactic ambiguity resolution while reading in second and native languages. Quarterly Journal of Experimental Psychology, 50, 119–148.

Kratzer, A. (1996). Severing the external argument from the verb. In Roryck, J. and Zaring, eds. Phrase Structure and the Lexicon. Dordrecht: Kluver. 109-138.

Mauner, G. & Koenig, J. P. (1999). Lexical Encoding of Event Participant Information. Brain and Language 68, 178– 184.

Extractions from relative clauses, a type of long-distance filler-gap dependency (FGD), typically yield unacceptable sentences across the majority of languages. Noun phrases involving relative clauses are therefore assumed to universally comprise syntactic “islands” for extraction (Ross 1967). The fact that extractions from relative clauses (RCEs) (1) are judged as acceptable in Swedish (Erteschik-Shir 1973, Engdahl & Ejerhed 1982) is thus unexpected. None of the theoretical accounts have proven satisfactory. Furthermore, no online, processing-based studies have investigated these structures.

Our study uses an eyetracking while reading paradigm to determine (i) whether RCEs elicit similar processing costs as extractions from non-restrictive relative clauses (nRCE) (2), which show island-like behavior in Swedish (Engdahl, 1997), or if they pattern closer to nonproblematic FGD sentences in which an extraction has been made from a that-clause (TCE) (3), and (ii) whether non-structural factors (frequency, pragmatic fit, and working memory) contribute to any pattern of effects, as facilitatory effects would serve as a positive heuristic for the non-island status of Swedish RCEs (Sprouse, Wagers, & Phillips, 2012; Traxler & Pickering 1996). An intransitive control condition involving pseudo-coordination within a relative clause (4) was also included.

(1) Såna där gamla skottkärror såg jag en man som alltid tvättade på macken när...

such there old wheelbarrows saw I a man that always washed at gas-station-the when...

(2) Såna där gamla skottkärror såg jag en man som förresten tvättade på macken när …

such there old wheelbarrows saw I a man that by-the-way washed at gas-station-the when…

(3) Såna där gamla skottkärror såg jag att en man alltid tvättade på macken när...

such there old wheel barrows saw I that a man always washed at gas-station-the when...

(4) Såna där gamla skottkärror såg jag en man som alltid stod och tvättade på macken…

such there old wheelbarrows saw I a man that always stood and washed at gas-station-the...

For our experiment, we tested 80 critical items, each with four structural variants (1-4) rotated across four presentation lists. Sixty fillers were also included. Forty-five participants each read one list and then completed two working memory span (WM) tests. We used linear mixed models to analyze first fixation durations, gaze durations, regression path durations, and total dwell times for two regions (embedded verb: tvättade; PP: på macken). To assess the contribution of non-linguistic factors, two measures of WM, Ospan (OS) and Dspan (DS); the pragmatic fit (Prag) of the filler to the embedded verb; and the frequency by which the embedded verb is followed by the filler NP were included as predictors. The primary finding at the embedded verb was that RCE and TCE patterned faster than nRCE in both early measures and in regression path durations, with RCE’s facilitation often being enhanced by or conditional upon increases in OS and Prag. This suggests that RCE is processed more similarly to TCE than to nRCE when verb/object integration first occurs, as a function of non-structural factors. Total Durations at the verb region exhibited a three-way distinction, in which RCE patterned between nRCE and TCE (which showed the greatest facilitation), signaling that integrative processes may be somewhat more difficult for RCE than TCE over time, but are still easier for RCE than for nRCE. At the PP region, both RCE and TCE patterned together faster than nRCE, though for RCE this facilitation was again often dependent on non-structural factors, emerging (for Total Durations) and becoming stronger (Gaze Durations) only at higher values of Prag and Ospan.

Our findings provide evidence that RCEs are easier to process than nRCEs, and that this facilitation is dependent in part on high values of certain non-structural factors, such as working memory span and the pragmatic fit of the filler to the selecting verb. Our study thus provides novel processing evidence that Swedish RCEs are more appropriately categorized as long-distance FGDs and not as syntactic islands.

Relative clauses are considered strong islands for extraction across languages. Swedish comprises a well-known exception, allegedly allowing extraction from relative clauses (RCE), raising the possibility that island constraints may be subject to "deep variation" between languages. One alternative is that such exceptions are only illusory and represent "surface variation" attributable to independently motivated syntactic properties. Yet, to date, no surface account has proven tenable for Swedish RCEs. The present study uses eyetracking while reading to test whether the apparent acceptability of Swedish RCEs has any processing correlates at the point of filler integration compared to uncontroversial strong island violations. Experiment 1 tests RCE against licit that-clause extraction (TCE), illicit extraction from a non-restrictive RC island (NRCE), and an intransitive control. For this, RCE was found to pattern similarly to TCE at the point of integration in early measures, but between TCE and NRCE in total durations. Experiment 2 uses RCE and extraction from a subject NP island (SRCE) to test the hypothesis that only non-islands will show effects of implausible filler-verb dependencies. RCE showed sensitivity to the plausibility manipulation across measures at the first potential point of filler integration, whereas such effec.ts were limited to late measures for SRCE. In addition, structural facilitation was seen across measures for RCE relative to SRCE. We propose that our results are compatible with RCEs being licit weak island extractions in Swedish, and that the overall picture speaks in favor of a surface rather than a deep variation approach to the lack of island effects in Swedish RCEs

Noun phrases involving relative clauses are assumed to universally comprise syntactic “islands” for extraction, but Swedish exists as a possible exception. Using eyetracking while reading, we investigated whether extraction from Swedish restrictive relative clauses (RCE) ([such old wheelbarrows]1 saw I a man that always washed __1 with benzine...) elicit similar processing costs as extractions from non-restrictive relative clauses, which are known to comprise strong islands (StrongIs); or if they pattern closer to extractions from non-island constructions (NonIs). We also examined to what extent non-linguistic variables (working memory WM, verb-object frequency, and pragmatic-fit) contribute to such differences. Results from a mixed models analysis of the embedded verb (washed) and spillover region (with...) suggest that in early measures, both RCE and NonIs show facilitation relative to StrongIs, but in late measures, RCE patterns closer to StrongIs as WM and pragmatic-fit increase, suggesting that Swedish RCE acceptability is partly dependent on non-linguistic factors.

The current study compares controlled acceptability judgments for Swedish relative clause extractions to extractions from <i>that</i>-clauses and extractions from non-restrictive relative clauses. It also compares each structure in both extracted and non-extracted form. The reported possibility of relative clause extraction in Swedish (and the other Mainland Scandinavian languages) has long presented a challenge to universal theories of constraints on extraction because the phenomenon is cross- linguistically very rare. In the off-line judgment data presented here, relative clause extractions are shown to pattern with extractions that are assumed to involve an island-like violation (non-restrictive relative clause extraction), thus contrasting with informal judgments reported in the literature. The data also appear to present a counterpoint to the conclusion reached in Tutunjian, Heinat, Klingvall and Wiklund (2017) from on-line eye-tracking measures, regarding the representational status of this structure as being more in line with that of a licit extraction. Potential explanations for the obtained patterns of the result are discussed, presenting avenues for further investigations.

Purpose - This study explores variation theory-based interventions' contribution to enhancing preschoolers' subitizing abilities during iterative lessons. Results are presented according to low, middle and high achievers, focused on those with prominent challenges. Design/methodology/approach - In total 68 Swedish preschoolers participated. They were randomly assigned to one of the intervention/control groups: 23 mixed design, 24 variation theory design and 21 controls. Data from 59 pupils were analyzed. A learning study with three lesson designs for each intervention group was used in autumn 2022 and spring 2023, respectively. The mixed design met ordinary textbook material and variation theory (VT) group instructions designed by VT during autumn, whereas all met instructions designed by VT in spring. Each group was divided into two subgroups. The controls followed business-as- usual math teaching. Two tests assessed pupils' results before and after the interventions. Findings - All groups showed significant improvements between pre- and post-tests. Pupils in the mixed and VT groups developed more knowledge than the controls. The low-achieving group developed the most. The VT group exhibited a more complex pattern of variation and a greater increase than the mixed group. The intervention groups showed a greater improvement in both tests' results than the controls. Originality/value - There is agreement on the importance of subitizing abilities as a predictor of future mathematical development; however, lesson design and research on pupils' development regarding lessons offered are limited.