Konrad Szcześniak University of Silesia in Katowice, Poland Benefits of L1–L3 Similarities The Case of the Dative Case A b s t r a c t The present study looks at the effects of structural similarities between L1 and L3 on learner performance. The main focus is on the knowledge of the dative construction in Polish learners of Portuguese. Portuguese and Polish share a number of important similarities in their use of the dative, which include functions not found in English, the learners’ L2. The main question investigated in this study is whether learners are aware of the non-banal similarities (given that they do not occur in English) and whether they can make use of that knowledge. It is found that learners do benefit from positive transfer, but that depends on the kind of function and frequency of the dative. Keywords: dative, transfer, avoidance, priming Introduction The role of L1 in the learning of a foreign language has been an important recurring theme since the beginning of second language acquisition research. It has also received conflicting interpretations, ranging from proposals that transfer from L1 is of negligible importance (e.g., George, 1972), through the recognition that it does affect performance in a foreign language, usually nega- tively (Cornips, 1992), to more nuanced analyses showing that much depends on proficiency. Thus, recent studies suggest that “as proficiency increases meaning-related transfer becomes more common” (Agustín Llach, 2010, p. 6), while “form-focused transfer is most dominant in the early stages of language learning” (Gabryś-Barker, 2006, p. 145). Theory and Practice of Second Language Acquisition vol. 3 (2) 2017, pp. 87–110 Konrad Szcześniak88 In this study, I wish to focus on the role of structural similarities between an L1 and L3 in acquisition and performance. Specifically, I will look at ele- ments of grammar that are shared by the two languages, but are at the same time fairly remarkable in the sense that they are not found in the learner’s L2. This will make it possible to correlate very concrete linguistic data with performance data and revisit the long-standing debate on whether contrastive analysis is relevant to learning. I will question the view that “[i]nterference, or native to target language transfer, plays such a small role in language learn- ing performance that no contrastive analysis, no matter how well conceived, could correlate highly with performance data, at least on the level of syntax” (Whitmann & Jackson, 1972, p. 40), and I will argue that similarities uncovered by comparison can in fact strongly facilitate learning. To explore the question of positive transfer, the present study examines Polish learners’ perception of the dative construction in Portuguese. Similar studies investigated the acquisi- tion of the dative construction in a foreign language, although to the best of my knowledge none featured the pair Polish-Portuguese. In one, Zyzik (2006) reports on the acquisition of the Spanish dative by English-speaking learners, and while the findings of that study are relevant to our discussion, the combi- nation of languages it involved determined a focus quite unlike in the present study. That is, because there are few common uses of the dative in Spanish and English, learners cannot fall back on their L1 knowledge to assist consolidation of the L2 system, so Zyzik’s study did not explore the benefits of similarities. On the other hand, the present investigation looks at how certain uses of the dative construction, common to L1 and L3, can be consolidated thanks to the similarities. In the following section, a description of the dative is offered, fo- cusing mainly on similarities between Portuguese and Polish, especially those points that set the two languages apart from English. The Dative The Dative Alternation Indo-European languages are characterized by the presence of the dative case, which marks indirect objects, typically expressing recipients in situations of transfer of possession (as in: Pass me the salt). Even in languages without overt case, like English, the dative case affects the behavior of various gram- matical constructions. For example, those verbs that alternate between the dative and prepositional frames are sensitive to the nature of the entity candidate for dative case marking. While animate entities can freely receive dative case and Benefits of L1–L3 Similarities. The Case of the Dative Case 89 thus alternate freely between the prepositional and dative frame (1), inanimate participants resist the alternation (2). The logic of the constraint blocking uses like (2b) is that since the dative case prototypically marks recipients and po- tential possessors, inanimate entities incapable of possession do not qualify for dative marking. (1) a. Jordan threw the ball to Carter. (Prepositional frame) b. Jordan threw Carter the ball. (Dative frame) (2) a. Jordan threw the ball to the basket. (Prepositional frame) b. *Jordan threw the basket the ball. (Dative frame) According to Gropen et al. (l989) and Pinker (1989), the constraint on animacy is universal, and as such can be found in many languages. Indeed, such is the case of other European languages, where inanimate participants can only be expressed through a goal to-pp, but not a dative frame. In Portuguese and Polish, such uses are thus also anomalous. The following is a Polish translation of example (2): (3) a. Jordan rzucił piłkę do kosza. Jordan threw ball to basket ‘Jordan threw the ball to the basket.’ b. *Jordan rzucił oszowi piłkę. Jordan threw basket-dat ball ‘Jordan threw the basket the ball.’ This shows that in their exposure to the dative system in a new language, learners should be able to rely on at least some principles operative in the functioning of dative objects that are not language-specific, but indeed com- mon to more languages. Still, there are particularities in the way the dative construction is used. Some uses are common to L1 and L3, and they are especially interesting when they do not occur in L2. They will be discussed below. Dative of Possession Of great interest is the range of meanings which are expressed by the indi- rect complement construction. Portuguese makes use of indirect complements to indicate a possessor affected in an event (dativo de posse, Bechara, 2002, p. 425; Rocha Lima, 2011, p. 308). Konrad Szcześniak90 (4) Lavaram-lhe o pescoço e cortaram-lhe o cabelo. (They) washed-him the neck and cut-him the hair. (Jorge Schwartz, Vanguardas argentinas) ‘They washed his neck and cut his hair.’ (5) […] curar-lhe a filha. (Raimundo Lúlio, O Livro das Bestas) cure-him the daughter ‘to cure his daughter’ (6) Partiram-lhe os vidros e roubaram-lhe o rádio. (They) broke-him the windows and stole-him the radio (Joana Vilela, Lisboa, anos 70) ‘They broke his windows and stole his radio.’ These uses are strikingly similar to the behavior of the Polish dative, which is also typically employed to express the affected possessor. The above sentences can be translated almost word for word, save the inexistence of articles in Polish: (7) Umyli mu szyję i obcięli mu włosy. (They) washed him neck and cut him hair. ‘They washed his neck and cut his hair.’ (8) […] uleczyć mu córkę. cure him daughter. ‘to cure his daughter’ (9) Wybili mu szyby i ukradli mu radio. (They) broke him windows and stole him radio. ‘They broke his windows and stole his radio.’ The similarity between Polish and Portuguese can better be appreciated when they are contrasted with English. In the construction which describes an activ- ity directed toward an object and its possessor, Polish and Portuguese mark the affected possessor by means of the dative pronoun, whereas English indicates it by means of the possessive pronoun: Table 1 Form of dative of possession constructions in Polish, Portuguese, and English Polish Portuguese English Cut him [hair] Obetnij mu [włosy] Cut him [the hair] Corta-lhe [o cabelo] Cut [his hair] Benefits of L1–L3 Similarities. The Case of the Dative Case 91 Ethical Dative An even more remarkable similarity can be observed in the use of the so- called ethical dative (dativo ético, Bechara, 2002, p. 425). Here, a participant is mentioned that is not directly involved in the situation (Svobodová, 2014, p. 54)—it is not an argument of the verb, but is mentioned as an “affected party,” a participant emotionally involved in the situation. The participant in question is usually the speaker. In some cases, the participant can be rendered in English as a colloquial dative (e.g., sing me a song); in other cases it is most naturally not mentioned at all (11); and in still other situations it is expressed by the malefactive on construction (12). (10) Abre-me essa porta. (José Leite de Vasconcellos, Revista Lusitana, 1917) Open-me that door ‘Open that door.’ (11) Não me enviem cartões a essas pessoas. (Bechara, 2002, p. 425)  Not me send postcards to those people  ‘Don’t send postcards to those people.’ (12) Não me chores agora. (Mário de Carvalho, Apuros de um pessimista  Not me cry now  em fuga, 1999)  ‘Don’t cry on me now.’ Predictably, these uses are also available in Polish (13)–(15). (13) Otwórz mi te drzwi.   Open me that door   ‘Open that door.’ (14) Nie wysyłajcie mi kartek do tych osób.   Not send me postcards to these people   ‘Don’t send postcards to those people.’ (15) Nie płacz mi teraz.  Not cry me now  ‘Don’t cry on me now.’ It should be pointed out, however, that this special use of the dative construction is not very common in either Portuguese or Polish. It is typically only found in spoken language, and is used on special occasions, when the speaker feels Konrad Szcześniak92 the need to stress his or her emotional involvement toward the outcome of the situation being discussed. One can speculate that a speaker may reasonably be imagined as functioning successfully in either language, without using or hearing this element for extended periods of time. Dative of Opinion One point of dissimilarity between Polish and Portuguese is the use of the dative case to express a scenario including its participant’s perspective. (16) É-lhe claro que existe qualquer coisa de maior.   Is-him clear that exists some thing of greater   (http://www.revistapassos.pt, accessed May 12, 2017)   ‘It is clear to him that there exists something greater.’ (17) É-nos proveitoso neste mundo utar.   Is-us advantageous in-this world fight   (www.ocalvinista.com, accessed May 12, 2017)   ‘It is worth our while to fight in this world.’ (18) Não me é possível traduzir em palavras o que sinto.  Not me is possible translate in words the what (I) feel   (Juscelino Kubitschek, Discurso de 1960)  ‘It is not possible for me to translate into words what I feel.’ It should be noted that in Polish too some meanings can be expressed by means of this dative construction (19)–(20). In some cases, usage similarities are quite striking, so much so that the familiar classic line humani nihil a me alienum puto is rendered through essentially the same dative structure in both Polish and Portuguese (21). (19) Jest mi obojętne, jakiej są płci. (Mirosław Bujko, Złoty pociąg)   Is me indifferent what are sex   ‘It doesn’t make any difference to me what sex they are.’ (20) Nie było mu trudno takie rodziny znaleźć.   Not was him hard such families find   (Kacper Zagadka, O Małym Królu)   ‘It wasn’t hard for him to find such families.’ Benefits of L1–L3 Similarities. The Case of the Dative Case 93 (21) a. Sou homem; nada humano me é estranho. (I) am man; nothing human me is strange ‘I am a human being; I consider nothing that is human alien to me.’ b. Jestem człowiekiem; nic co ludzkie nie jest mi obce. (I) am man; nothing that human not is me strange ‘I am a human being; I consider nothing that is human alien to me.’ However, while Polish has a rather limited selection of predicative adjectives that take dative complements, in Portuguese, the construction allows any ad- jective that expresses the subject’s opinion toward the proposition presented in the sentence. Below is a brief sample of adjectives attested in the construction: (22) a. Era-lhes óbvio que... (‘it was obvious to them that…’)   b. Era-lhe estranho que... (‘it was strange to him that…’)  c. É-me duvidoso que… (‘it is doubtful to me that…’)  d. É-me pouco relevante que… (‘it is not very relevant for me that…’)  e. Essa questão é-me profundamente indiferente… (‘the question is profoundly indifferent to me…’)  f. É-me assustador sequer pensar… (‘I find it frightening to even think…’)  g. É-me evidente que... (‘It is evident to me that…’) The similarities between Polish and Portuguese, relative to English are summed up in Table 2. Table 2 Presence or absence of three functions of the dative in Portuguese, Polish, and English Dative of possession Ethical dative Dative of opinion Portuguese + + + Polish + + –/+ (only single adjectives) English – – – Konrad Szcześniak94 Learnability Challenge The above uses of the Portuguese dative construction present a challenge for learnability theories. The most obvious question is whether they are mastered by Polish learners of Portuguese as a third language,1 and if so, whether their successful usage can be attributed to similarities with Polish. As experienced language learners, they can be expected to be aware that uses like the ethical dative or the dative of possession are far from universally available options, since they are not found in English. Their existence in Polish and correspond- ing absence in English may prompt learners of Portuguese to employ avoidance strategies on the grounds that it would be beyond belief that such evidently special L1 features could also be found in another language. Even when they have witnessed similar uses, Polish learners of Portuguese may still avoid them in their own production, because “when great similarities exist […] the learner may doubt that these similarities are real” (Gass & Selinker, 1994, p. 138). On the other hand, one can also view learners’ perception of similarities more optimistically. It is equally possible that once they are noticed, such simi- larities may be taken advantage of in building proficiency in a L3. Such a turn of events would be consistent with the idea that “[t]he use of prior knowledge, i.e. of one’s L1, could contribute to learners’ success […] at the cognitive level (objective language progress)” (Gabryś-Barker, 2006, p. 164). One can expect such success to be contingent on the learners’ level and prior knowledge that new competences can be built upon. This is consistent with the Teachability Hypothesis, which claims that […] the acquisition process cannot be steered or modelled just according to the requirements or precepts of formal instruction. On the contrary, […] teaching itself is subject to some of the constraints which determine the course of natural acquisition. (Pienemann, 1989, p. 57). Points of an L2 grammar can only be acquired successfully when the learner is ready for them; that is, their acquisition should be preceded by first master- ing those points of grammar that are earlier and more basic in the acquisitional chronology. This presupposes a natural sequence in which the uses of dative outlined in the Section The Dative are arranged. Thus, it is fair to consider typical transfer-of-possession uses (Eu dei-lhe flores, ‘I gave her flowers’) as being the first stage in the acquisition of the dative construction. This is then 1 It is assumed here that for Polish learners in general, Portuguese almost always is a third language. When Polish speakers become interested in learning Portuguese, they already know at least English and in many cases, they are also f luent in Spanish, as is the case with the subjects in the present study. Benefits of L1–L3 Similarities. The Case of the Dative Case 95 followed by uses where the ‘possession’ element is present but is secondary to the construction’s focus on affectedness (Dative of possession and Ethical da- tive), which are in turn followed by uses where possession is most backgrounded and is not commonly found in L1 (Dative of opinion). Similar conclusions about the existence of a sequence of acquisition of the dative system are presented in Zyzik (2006, p. 132). If transfer-of-possession uses have been internalized, the learner should be ready for the Dative of possession and Ethical dative, especially if these two applications of the dative are also present in L1. And there is good reason to believe that that is in fact what happens. It has been demonstrated that language users are sensitive to unusual patterns in the input, which are incorporated into a speaker’s linguistic knowledge as a result of mere exposure, as argued by authors like Bybee (2010) and Taylor (2012) who hypothesize that “each linguistic encounter lays down a trace in memory” (Taylor, 2012, p. 3). Specifically, what seems to happen is that as language users attend to input, certain features present in that input register in their minds as salient points likely to have a long-lived effect on their linguistic knowledge. It should be pointed out in this connection that salience and noticeability have obvious implications for second language acquisition. Under the Noticing Hypothesis (Schmidt, 1990), it is a measure of successful learners (of a foreign language) to what extent they are able to notice new elements of the input that they were not previously aware of and to go on to use these forms themselves. It is important to stress that the new features need not even be noticed consciously. A large part of knowledge is built without conscious report, as is indicated by the way people reuse elements of preceding discourse in their own production. For example, Levelt & Kelter (1982) show that the choice of grammatical constructions used when answering a question depends on the very question’s construction. If a person is asked At what time do you close? the answer is very likely to include the preposition at, but the preposition is equally likely to be absent in the answer if the question lacks it (What time do you close?). While speakers are clearly sensitive to the “question’s surface form [which] can affect the format of the answer given” (Levelt & Kelter, 1982, p. 78), most probably, they are not even aware that their answer echoed the form of the question. What such findings suggest is that the presence of an element in the input is powerful enough to prime the learners to echo that element in their own use, and the strength of priming depends on the salience and noticeability of a given element. The Portuguese dative is an interesting case in the context of salience-driven priming and use, because its salience should be heightened by its remarkable similarity with Polish. It is evident that such features do get noticed, for it would be rather counterintuitive to view exposure to input as having no effect on the learner’s knowledge. It would be interesting to see how the “trace in memory” envisioned by Taylor (2012) translates into actual use. Konrad Szcześniak96 Thus, the question to be explored is what effect on the learner the similarities with L1 have. Are they dismissed as “suspicious,” “too good to be true” and therefore worthy of avoidance, or, on the contrary, do they contribute to prim- ing and, in consequence, to more felicitous learning? Study Research Questions The study aims to address the following questions: 1. Can L1 knowledge be tapped in developing an L3 system in areas of great similarity? Can such similarities neutralize interference from L2? One fairly peculiar use of the dative construction investigated in this study is governed by identical L1 and L3 rules, and it is therefore of great interest whether learners can transfer the construction’s specifications from L1 to L3. 2. Does recent exposure to uses of a construction prime adequate use of that construction? In other words, can the learner benefit from encountering a specific use of the construction without explicit instruction regarding its function? 3. Are the functions of the dative construction discussed in sections: Dative of Possession, Ethical Dative, and Dative of Opinion located within a natu- ral hierarchy determining their chronology of acquisition, in line with the Teachability Hypothesis? Participants The study involved 25 subjects, third-year students of Portuguese, aged 20–23, enrolled in the Spanish-Portuguese program at the University of Silesia in Katowice. At the time of the study, the subjects had completed three years of intensive training in Portuguese, attending various classes in Portuguese as the language of instruction. The group included 19 females and six males. All subjects were native speakers of Polish, with English being their L2 and Portuguese coming as an L3. None of the subjects reported having visited a Portuguese-speaking country before. Benefits of L1–L3 Similarities. The Case of the Dative Case 97 Task The subjects were instructed to complete ten sentences with blanks and multiple-choice options. The blanks referred to verb complements, of which six included dative objects and these were the focus of the study. The remaining four sentences contained accusative and other complements and were included to make it harder for the subjects to discover the objective of the study (which could prompt them to select the correct answers for the wrong reasons). The sentences presented to the subjects had the following format (The sheet with all sentences is shown in Appendix C). The sentence in Figure 1 can be translated as ‘Paulo says that they broke the windows in his car and stole the radio,’ and the correct answer is C, which involves a dative complement, literally ‘broke him the windows.’ 6. O Paulo diz que --------- partiram os vidros no carro e roubaram o rádio. A. ele C. lhe B. o      D. dela Figure 1. Format of sentences used in the task. The six sentences with dative complements included two that exemplified the ethical dative, two with the dative of possession, and two with the dative of opinion. At this point it is necessary to address one possible objection, namely the possibility that filling blanks may not represent the most natural demonstra- tion of performance. Indeed, apart from completing forms or questionnaires, natural use of a language in communicative settings rarely involves tasks typical of classroom work. Nevertheless, the format adopted in the study can be considered a reliable tool measuring the learner’s level of acquisition of the system. To complete the sentences, it is necessary to understand the content of the sentence, identify the semantic element of possession and associate it with the correct use of the dative construction. This may not guarantee active use in completely free-choice expression, but it does show the learner’s growing understanding, passive or active, of the dative functions. The subjects were divided into two groups, with 13 and 12 participants in each. One group was first asked to read three short paragraphs in Portuguese which featured the three uses of the dative in Portuguese. The idea was to measure possible effects, if any, of structural priming: Does exposure to a given construction affect performance in the use of that construction with other verbs? To ensure that the subjects would not mechanically skim over the relevant uses of the dative, they were instructed to answer comprehension questions that required focusing on the information conveyed by means of the dative constructions (see Figures 2 and 3). Konrad Szcześniak98 É-me indispensável estar perto do mar, da Natureza, ouvir o silêncio, e Portugal tem paisagens magníficas; impossível não regressar com energias renovadas. Fonte http://caras.sapo.pt/famosos/2010-04-02-a-escolha- de...-sofia-nicholson 2. A autora precisa da Natureza para A. escrever sobre ela. B. tirar fotografias. C. nadar no mar. D. descansar. Figure 2. A comprehension task used to prime the subjects. It is indispensable for me (lit. it is me indispensable) to be close to the sea, to Nature, hear the silence, and Portugal has magnificent landscapes; impossible to not come back with renewed energy. Source http://caras.sapo.pt/famosos/2010-04-02-a- escolha-de...-sofia-nicholson 2. The author needs Nature to A. write about it. B. take pictures. C. swim in the sea. D. rest. Figure 3. Translation of the example shown in Figure 1. Three short paragraphs were also shown to the subjects in the other group (control), but the passages in this group did not include uses of the dative construction. The objective of using non-dative texts in the control group was to observe differences in the effect on performance in the case of priming and the absence of priming. The sheets with paragraphs and comprehension tasks are given in Appendix A (group with priming) and Appendix B (control group, no priming). Results The first salient observation is a clear bias in the distribution of the correct answers among the three functions of the dative construction. The best perform- ance is found in the use of the dative of possession (100%, 100% in group A, and 92%, 77% in B); followed by the dative of opinion (42%, 25% in A; 23%, 8% in B) and the ethical dative (8%, 0% in A; 15%, 0% in B). The uses of the dative of possession were recognized correctly by almost all subjects, whereas the ethical dative sentences were missed by a great majority of subjects, with the dative of opinion sentences faring only slightly better. It should be noted that the contrast is so strong that it cannot be a result of coincidence or ran- dom fluctuation. For example, in the primed group, 12 correct answers were given to each sentence exemplifying the dative of possession (i.e., all subjects got them right) and only one correct answer to the ethical dative examples (with 23 incorrect answers). Thus, the difference between the correct answers given to the dative of possession examples and the ethical dative examples is statistically significant, with its Two Tailed p-value and One Tailed p-value both Benefits of L1–L3 Similarities. The Case of the Dative Case 99 at 0, which means there is a 100% chance the proportions are different. The percentages of correct answers are shown in Figures 4 and 5. Figure 4. Correct choices (in %) for the sentences with the dative of posses- sion (1–2), the dative of opinion (3–4), and the ethical dative in group A (with priming). Figure 5. Correct choices (in %) for the sentences with the dative of possession (1–2), the dative of opinion (3–4), and the ethical dative in group B (no priming). Priming does seem to play a role, but it is stronger in the case of the dative of possession. As Figures 4 and 5 illustrate, the percentage of correct answers (the first two columns) in the unprimed control group, although high, is visibly lower than in the group that benefited from priming. Although the difference Konrad Szcześniak100 may not be statistically very significant in the case of the first sentence (with only a 52% chance that the proportions are different), the difference is much more pronounced in the second example, where there is a 91.8% chance that the proportions are different. In the first pair, the two-tailed p-value is 0.48 and the one-tailed p-value: 0.26; in the second pair, the two-tailed p-value is 0.08, and the one-tailed p-value: 0.04. In the case of the dative of opinion (examples 3–4), the chances that the differences between the results in the primed and the unprimed group are significant are 67% and 75.2%, and in the ethical dative sentences (5–6), the chances are 40% and 0%. The diminishing trend in the effects of priming can be seen in the orientation of the dotted line in Figure 6. Figure 6. Probabilities (in %) that the proportions between primed and unprimed groups are different. Discussion The examples tested did not involve typical uses of the dative case, ones that can be found in English (i.e., transfer of possession uses discussed in the Section The Dative Alternation), so the subjects could not rely on the “uni- versal” logic of the dative construction. To give correct answers, the subjects had to tap their knowledge of more language-specific properties of the dative construction described in the following sections: Dative of Possession, Ethical Dative, and Dative of Opinion. Thus, the likeliest explanation of the data presented here is transfer from L1, or at least lack of strong L2 transfer. Because the dative of possession is fairly common in Polish, it can be assumed that when it is noticed in Portuguese, it is internalized most easily. Note that almost all the subjects recognized and Benefits of L1–L3 Similarities. The Case of the Dative Case 101 applied correctly the use of the dative of possession in both sentences that exemplified it, in both the primed and unprimed group. To appreciate the sig- nificance of the data obtained, it is necessary to consider the two sentences in some detail. They featured two kinds of possession, namely alienable (23) and inalienable possession (24). (23) O Paulo diz que lhe partiram os vidros no carro e roubaram o rádio.   Art Paul says that him-dat broke art windows in-art car and took art radio   ‘Paulo says that they broke the windows in his car and took the radio.’ (24) A Antónia acredita que foi a viagem a Fátima que lhe curou o corpo.   Art Antónia believes that was art trip to Fátima that her healed art body   ‘Antónia believes that it was the trip to Fátima that healed her body.’ In the example with alienable possession, the radio is seen as being possessed more “distantly”; that is, it is not an inalienable part of the subject’s body. In example (24), on the other hand, the body is a typical instance of inalienable possessum, more intuitively associated with the grammatical category of pos- session. The decision to include two such different kinds of possession in the sentences test was motivated by the need to avoid skewing the answers given to some types of the dative construction. That is, one would expect better per- formance in the case of the dative of possession relative to the other uses of the construction, if the examples used in the form were disproportionately more transparent or in any way more accessible than those representing the other uses. The inclusion of the alienable possession sentence (23) was designed to make this part of the task closer to the other uses of the dative construction in terms of difficulty. In the case of the dative of possession sentences, the alien- able possession use could be considered more challenging, and indeed this is reflected in the results which are lower than for the inalienable possessive use, but only slightly. What needs to be underscored is that the sentence was never- theless handled much better than the other uses. That is, one can conclude that the difference in performance is not due to the objective, informational content of the sentences, but due to grammatical difficulty. The dative of possession is more accessible because it is a deeply entrenched part of Polish grammar, unlike the other two uses, which although present, are not as frequent. The dative of opinion and ethical dative contrast strongly with the possessive use of the dative construction in that they are considerably less common in Polish. While the ethical dative functions the same way in both languages, it is far from being a basic feature of the grammar of either language. And the dative of opinion, although attested in Polish, is used with a very restricted range of adjectives. As a result, learners of Portuguese have limited chances of coming across instances of these functions with sufficient frequency for the construc- Konrad Szcześniak102 tion to be noticed and internalized. When uses like the ethical dative do get noticed only sporadically, their appearance in Portuguese may strike learners as either implausible or downright incomprehensible, and consequently these functions may not be recognized as such at all. The differing effects of priming for the dative of possession on the one hand, and the ethical and opinion dative on the other can be accounted for in terms of their frequency, too. It could be speculated that the uses of possession dative shown to the subjects before the task helped pre-activate the function, because it was already present in their minds in one form or another. Although some effect could be observed, priming failed to translate significantly into better performance in the case of the other two datives, probably because there was little to prime: the structures were not part of the subjects’ knowledge of Portuguese, so they were missed in the priming input, or the right association was not made when it came time to use the knowledge in the multiple-choice task. All this justifies the conclusion that transfer from L1 is contingent on a number of factors. One is the learners’ proficiency level. The subjects exam- ined in the study were intermediate learners with enough experience to take advantage of the observed similarities through positive transfer from L1. This represents an evident improvement over the tendency to transfer from L2, often negatively, at lower levels of proficiency in L3 (Angelovska & Hahn, 2012). For example, in the case of beginning learners of Portuguese, negative transfer from L2 is the only explanation for the use of the verb ser (‘to be’) in the “age description” collocation. Polish students tend to build sentences like *Eu sou 20 anos (‘I am 20 years’) instead of Tenho 20 anos (‘I have 20 years’) because their L2 English interferes, even though transfer from their L1 Polish would have been yielded the right structure. At more advanced stages, transfer tends to “even out,” with both L1 and L2 serving as sources for more positive transfer. The results of the study also justify the observation that, to the extent that the subjects’ correct responses can be attributed to transfer from L1, the effects of transfer outweigh avoidance mainly in the case of more common construc- tions. It seems that the reason why the subjects’ performance was more reliable in the case of the dative of possession than in the ethical or opinion dative is that the former is a more common feature of grammar and therefore more likely to have been come across, and hence psychologically more entrenched. The latter two types are, by comparison, so rare that they may have been misidentified. It should be underscored that in the multiple-choice task, the subjects were at- tracted by distracters that were securely not keyable – the options selected were incorrect, being either ungrammatical or semantically anomalous or both, but it was these options that were nonetheless perceived as being more acceptable than the intended dative keys. This shows that the two rare functions of the dative were misinterpreted by the subjects. Benefits of L1–L3 Similarities. The Case of the Dative Case 103 Finally, the data demonstrate that the role of priming also depends on the frequency of the material. While the more frequent structures benefited from the subjects’ recent exposure to relevant examples in the input, their perform- ance on the low-frequency dative constructions barely changed at all in the primed group compared to the control. This finding can be viewed as being in conflict with the theory that priming effects increase as frequency decreases. That is, previous studies have shown that language users tend to notice and repeat those elements of input that are unusual or infrequent when compared to other forms. Thus, for example, observing the passive voice in the input can be trigger enough for the speakers to build a passive sentence themselves: even though the passive voice is not a particularly unusual form, it is considerably less frequent than the active voice, and is therefore likely to stand out in the input. This regularity has been named “the inverse frequency effect: the less frequent structure primes more” (Jaeger & Snider, 2008, p. 1064). By virtue of being less frequent, an expression has the so called “surprisal effect” which assures its noticeability. However, as the present study of the dative construction shows, the “inverse frequency effect” seems to cease when the frequency of a given construction is too low. As such, the finding is not really a contradic- tion of the inverse correlation of priming strength and frequency, but instead it complements it: the inverse correlation holds for most cases of “familiar” constructions such as the passive voice or the comparative degree of adjectives, and it correctly predicts that priming will be stronger for the less frequent member of a given pair of constructions. It can be hypothesized that language users do notice constructions that are relatively less frequent provided that these constructions are recognized and comprehended reliably and adequately. On the other hand, priming does not even begin to occur for forms that remain obscure to the learner because of their low frequency and are therefore not even tempting candidates for reuse after recent exposure. This sudden cessation of the correlation can be represented graphically by means of Figure 7. Figure 7. Inverse frequency effect with an abrupt drop for very infrequent forms. Konrad Szcześniak104 Outlook Unfortunately, there are limits to the strength of conclusions that can be drawn from the present data. As hinted above, correct answers in a multiple choice task do not guarantee correct use of the corresponding language points in the learners’ production. Indeed, there is no way of knowing how learners go about the use of the dative construction in their writing or speech, because as of today, no learner corpora of Portuguese output are available which could be searched for the use of specific language forms. Ideally, a research design for future investigation of factors such as priming, learners’ L1, or their level should make use of triangulated experimental and corpus data as corroborat- ing evidence. This, however, may remain a mere plan waiting for Portuguese learner corpora to be created. Learner corpora are available for English, such as the International Corpus of Learner English (ICLE, Granger et al., 2002), and such corpora were used to investigate the use of the dative construction by Polish and German learners of English (Callies & Szcześniak, 2008), but rare language combinations like Polish-Portuguese make the creation of similar tools rather unlikely any time soon. Another way to supplement judgment tasks like the tool used in the present study would be to use other data collection techniques such as elicited produc- tion, because experimental data are often needed to corroborate data obtained from one source (see Wasow & Arnold, 2005). Elicitation was not attempted here for reasons of space, but there is no theoretical reason why the learners’ knowledge of the applications of the dative should not be investigated through structured question-response tasks, where subjects are presented with pic- tures illustrating simple events and asked to describe who did what to whom. Admittedly, this would require a larger-scale project involving a more complex setup, with a substantial number of non-dative sentences included as controls, all of which would require a discussion twice the size of the present study. Nevertheless, such studies can be undertaken to verify the findings presented here. Conclusions The results of the present study confirm that positive transfer from L1 can override the influence of L2 in the acquisition of an L3 especially at higher levels of proficiency. Indeed, it would be counterintuitive, if not downright absurd, to find L1–L3 similarities being ignored or their knowledge not being put to work in acquisition. Especially in the case of relatively common con- Benefits of L1–L3 Similarities. The Case of the Dative Case 105 structions, learners seem to be aware of the correct form of L3 expressions, which correspond closely in form and function to their equivalents in L1, and then L2 appears to subside as a source of transfer. However, it is evident that avoidance, likely motivated by L2–L3 differences, does play a role when it comes to structures that are rare, albeit perfectly grammatical and natural in L1. It should be fairly uncontroversial to propose supplementing the classroom practice of such L3 forms with explicit instruction in order to demonstrate the similarities in question. Contrasting the learners’ L1 and L2/L3 and pointing out common properties can be beneficial insofar as it may help raise awareness of features that would otherwise remain obscure or misunderstood. These conclusions seem especially valid in light of the effects of priming. As the results show, priming from input to immediate use is stronger for those forms that the learners are capable of recognizing in the first place, which can be seen as a restatement of the i + 1 Hypothesis (Krashen, 1982). That is, priming occurred most visibly for forms within the learners’ reach, when they represented input within the i + 1 range. Conversely, priming decreased for rarer forms because they represented “input plus much more than one,” well beyond the learners’ competence level. To help learners benefit from priming (and transfer), their level would first need to be elevated, and one way of doing so is through explicit contrastive analysis of L1/L3. This is also consistent with the idea that, as predicted by the Teachability Hypothesis (Pienemann, 1989), successfully internalizing such finer points of grammar appears to be subject to constraints to do with the natural course of acquisition. It seems the more advanced functions of the dative construction cannot be mastered until the learner has cleared its most basic application, the transfer of possession. It is only after this most obvious and common use of the dative has been consolidated that the learner is ready for functions like the dative of possession or ethical dative, and these in turn precede the so-called dative of opinion, a function that comes and is learned last. References Agustín Llach, M. P. (2010). An overview of variables affecting lexical transfer in writing: A review study. International Journal of Linguistics, 2(1), 1–17. Angelovska, T., & Hahn, A. (2012). Written L3 (English): Transfer phenomena of L2 (German) lexical and syntactic properties. In D. Gabryś-Barker (Ed.), Cross-linguistic Influences in Multilingual Language Acquisition (pp. 23–40). Berlin–Heidelberg: Springer Verlag. Bechara, E. (2009). Moderna Gramática Portuguesa. Rio de Janeiro: Editora Nova Fronteira. Bybee, J. (2010). Language, usage and cognition. Cambridge: Cambridge University Press. Konrad Szcześniak106 Callies, M., & Szcześniak, K. (2008). Argument realisation, information status and syntactic weight – A learner-corpus study of the dative alternation. In M. Walter & P. Grommes (Eds.), Fortgeschrittene lernervarietäten korpuslinguistik und zweitsprachenerwerbsfor- schung (pp. 165–187). Tübingen: Niemeyer. Cornips, L. (1992). Syntactic interference in a bilingual community: The use of the ref lexive in intransitive variants of causative verbs. Intercultural Communication Studies, 2 (2): 23–41. Gabryś-Barker, D. (2006). The interaction of languages in the lexical search of multilingual language Users. In J. Arabski (Ed.), Cross-linguistic influences in the second language lexicon (pp. 144–166). Clevedon: Multilingual Matters. Gass, Susan M., & Selinker, L. (1994). Second language acquisition. New York: Routledge. Granger, S., Dagneaux, E., & Meunier, F. (2002). The international corpus of learner English. Handbook and CD-ROM. Lolvain-la-Neuve: Presses Universitaires de Louvain. George, H. V. (1972). Common errors in language learning. Rowley, MA: Newbury House. Gropen, J., Pinker, S., Hollander, M., Goldberg, R., & Wilson, R. (l989). The learnability and acquisition of the dative alternation in English. Language, 65, 203–257. Jaeger, F., & Snider, N. (2007). Implicit learning and syntactic persistence: Surprisal and cu- mulativity. In Brad. C. Love, Ken McRae, & Vladimir M. Sloutsky (Eds.), Proceedings of the 30th Annual Conference of the Cognitive Science Society (1061–1066). Washington, DC: Cognitive Science Society. Krashen, S. D. (1982). Principles and practice in second language acquisition. New York: Pergamon. Levelt, W. J. M., & Kelter, S. (1982). Surface form and memory in question answering. Cognitive Psychology, 14, 78–106. Pinker, S. (1989). Learnability and cognition. The acquisition of argument structure. Cambridge, MA: MIT Press. Rocha L., & Henrique C. (2011). Gramática Normativa da Língua Portuguesa. Rio de Janeiro: José Olympio Editora. Schmidt, R. (1990). The role of consciousness in second language learning. Applied Linguistics, 11, 129–158. Svobodová, I. (2014). Sintaxe da Língua Portuguesa. Brno: Masarykova Univerzita. Taylor, J. R. (2012). The mental corpus. How language is represented in the mind. Oxford: Oxford University Press. Wasow, T., & Arnold J. (2005). Intuitions in linguistic argumentation. Lingua, 1(15), 1481–1498. Zyzik, E. (2006 Learners’ overgeneralization of dative clitics to accusative contexts: Evidence for prototype effects in SLA. In C. A. Klee & T. L. Face (Eds.), Selected Proceedings of the 7th Conference on the Acquisition of Spanish and Portuguese as First and Second Languages (pp. 122–134). Somerville, MA: Cascadilla Proceedings Project. Konrad Szcześniak Die Nutzen aus den Ähnlichkeiten zwischen Erstsprache (L1) und Drittsprache (L3). Kasus: der Dativ Z u s a m m e n f a s s u n g Die vorliegende Arbeit betrifft strukturelle Ähnlichkeiten zwischen der Erstsprache (L1) und Drittsprache (L3) und deren Bedeutung für die im Fremdsprachenunterricht erzielten Benefits of L1–L3 Similarities. The Case of the Dative Case 107 Leistungen. Das Thema wird am Beispiel der Dativkonstruktion und deren Anwendung von den Portugiesisch lernenden Polen erörtert. Die polnische und die portugiesische Sprache kennzeichnet eine ganze Reihe von Ähnlichkeiten im Dativgebrauch, die solche Funktionen umfassen, welche in dem Englischen, das eine Zweitsprache für Portugiesisch Lernende ist, nicht vorkommen. Die Arbeit soll die Frage beantworten, ob sich die Fremdsprachenlernenden der bestehenden und nicht unbedeutenden Ähnlichkeiten (die im Englischen nicht auftreten) bewusst sind und ob diese Kenntnisse ihr Lernen begünstigen. Es wurde festgestellt, dass die Portugiesisch lernenden Polen im Stande sind, den positiven Transfer auszunutzen, das aber ist von konkreten Funktionen und von der Häufigkeit des Dativgebrauchs abhängig. Schlüsselwörter: Dativ, Transfer, Vermeiden, Bahnung Konrad Szcześniak108 A p p e n d i x A Leitura Costumes e comportamentos Leia os trechos e responda às perguntas. Claro que há coisas que me chateiam, como a toda a gente, mas penso sempre que tenho duas opções: posso deixar que isto me afete o suficiente para me pôr de mau humor e estragar-me o dia ou posso passar por cima, ir dar um passeio e pôr para trás das costas. Fonte http://caras.sapo.pt/famosos/2015-04-24-So- fia-Escobar-Este-ano-tem-sido-o-melhor-da- minha-vida 1. Podemos inferir que a autora do trecho evita que A. lhe estraguem o dia. B. muita gente pense mal dela. C. os seus amigos fiquem chateados. D. o passeio demore demasiado tempo. É-me indispensável estar perto do mar, da Natureza, ouvir o silêncio, e Portugal tem paisagens magníficas; impossível não regres- sar com energias renovadas. Fonte http://caras.sapo.pt/famosos/2010-04-02-a- escolha-de...-sofia-nicholson 2. A autora precisa da Natureza para A. escrever sobre ela. B. tirar fotografias. C. nadar no mar. D. descansar. Ele teve mãe – usamos a frase para definir homens cada vez mais raros. São os que, a qualquer hora, estão de barba feita e adequadamente vestidos. Parecem ter saído do banho naquele momento. Não só. São educados e gentis. Não poupam obrigados, desculpe-me e com licença. Dão prioridade à mulher, abrem-lhe a porta do carro e lhe pux- am a cadeira no restaurante. Convenhamos: o produto é cada vez mais raro, mas existe. Fonte htt p:// blogs.cor reiobraziliense.com.br/dad / filho_sem_mae/ 3. Segundo a autora, qual dos seguintes comportamentos não seria sinal de boa educação? A. Abrir a porta a uma mulher B. Oferecer banho a amigas C. Estar de barba feita D. Pedir desculpa Benefits of L1–L3 Similarities. The Case of the Dative Case 109 A p p e n d i x B Leitura Costumes e comportamentos Leia os trechos e responda às perguntas. Fria não consigo ser, porque vejo sempre à minha frente uma pessoa a lutar pelos seus sonhos e tenho imensa empatia com os con- correntes, mas objetiva sim, pois acredito que críticas construtivas nos ajudam a crescer e evoluir. Sou totalmente contra as críticas destrutivas. Fonte http://caras.sapo.pt/famosos/2015-04-24-So- fia-Escobar-Este-ano-tem-sido-o-melhor-da- minha-vida 1. Podemos inferir que a autora do trecho evita A. dar conselhos. B. falar com críticos. C. magoar as pessoas. D. lutar pelos seus sonhos. Gosto de conviver directamente com as pessoas, partilhar, viver outras realidades. Normalmente passo as férias de forma ativa, mas também para enriquecer-me pessoal- mente. Fonte http://caras.sapo.pt/famosos/2010-04-02-a- escolha-de...-sofia-nicholson 2. A autora provavelmente A. passa muito tempo na praia. B. fica em hotéis caros. C. fala com os locais. D. vê muita televisão. Seja respeitoso. Boa parte das pessoas se indigna com palavrões, obscenidades, lin- guagem vulgar e expressões chulas. Só as acolha em situações excepcionais. É o caso de manifestação de alguém quando a palavra tiver indiscutível valor informativo ou ref letir o seu estado emocional. Fonte htt p:// blogs.cor reiobraziliense.com.br/dad / filho_sem_mae/ 3. Segundo a autora, quando seria aceitável usar um palavrão? A. Ao pedir ajuda B. Ao falar com clientes C. Ao sentir dor muito forte D. Ao convidar amigos a uma festa Konrad Szcześniak110 A p p e n d i x C Preencha os espaços em branco usando uma das opções sugeridas. 1. E vocês nem -------- pensem em comer chocolate e gelado. Nada de açúcar! A. eu B. se C. me D. meu 2. O circo chegou -------. A. -lhe B. à cidade C. nem D. -o 3. Disse à polícia que não conhecia os suspeitos, nem ------- tinha visto. A. Ø B. suas C. os D. lhes 4. Não ------- voltes morto. Quero-te vivo. A. te B. o C. me D. lhe 5. Agora é ------- óbvio que votar à direita é uma melhor opção que votar à esquerda. A. mim B. -me C. meu D. -se 6. O Paulo diz que --------- partiram os vidros no carro e roubaram o rádio. A. ele B. o C. lhe D. dela 7. Depois de ouvir a palavra nova, o aluno escreveu --------- no caderno. A. sua B. deles C. -a D. -o 8. A Antónia acredita que foi a viagem a Fátima que --------- curou o corpo. A. seu B. dele C. a D. lhe 9. Quando a Ana recebeu a carta do Tiago, mandou --------- de volta. A. -no B. -lhe C. -a D. -o 10. Quando estamos no último andar do Empire State Building, é --------- assustador olhar para baixo. A. nós B. tanto C. -nos D. nosso