Top Banner
SSLA, 22, 535–556. Printed in the United States of America. MORPHOLOGICAL TYPE, SPATIAL REFERENCE, AND LANGUAGE TRANSFER Scott Jarvis Ohio University Terence Odlin Ohio State University This study clarifies issues related to the transferability of bound mor- phology and reports on an empirical investigation of morphological transfer in the spatial expressions of Finnish-speaking (n = 140) and Swedish-speaking (n = 70) adolescent learners of English. The re- sults indicate that both the bound, agglutinative morphology of the L1 Finnish spatial system and the free, prepositional morphology of the L1 Swedish spatial system constrain the types of options that learners pursue in their L2 English spatial reference. Additionally, however, the structural and semantic differences between the two L1 systems result in different patterns of spatial reference in the L2. We characterize these differences in terms of semantic transfer and simplification, and go on to show how transfer and simplification in- teract in our data. During the many years of interest in the phenomenon of language transfer, one of the most frequently discussed issues has been transferability. Again and again researchers have addressed the question of what constraints, if any, there are on cross-linguistic influence. Some have stated or implied that virtu- ally no constraints exist (e.g., Thomason & Kaufman, 1988), whereas others An earlier version of this paper was presented at the Second Language Research Forum, October, 1998. The authors would like to thank Kathleen Bardovi-Harlig, Donna Lardiere, Jussi Niemi, Bonnie Schwartz, Helena Sulkala, Sangeetha Venkatraman, and Don Winford for their help with different as- pects of this research. Address correspondence to: Scott Jarvis, Dept. of Linguistics, Ohio University, Gordy Hall 383, Athens, OH 45701; email: [email protected]. 2000 Cambridge University Press 0272-2631/00 $9.50 535
22
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Jarvis & Odlin 2000

SSLA, 22, 535–556. Printed in the United States of America.

MORPHOLOGICAL TYPE,SPATIAL REFERENCE, AND

LANGUAGE TRANSFER

Scott JarvisOhio University

Terence OdlinOhio State University

This study clarifies issues related to the transferability of bound mor-phology and reports on an empirical investigation of morphologicaltransfer in the spatial expressions of Finnish-speaking (n = 140) andSwedish-speaking (n = 70) adolescent learners of English. The re-sults indicate that both the bound, agglutinative morphology of theL1 Finnish spatial system and the free, prepositional morphology ofthe L1 Swedish spatial system constrain the types of options thatlearners pursue in their L2 English spatial reference. Additionally,however, the structural and semantic differences between the twoL1 systems result in different patterns of spatial reference in the L2.We characterize these differences in terms of semantic transfer andsimplification, and go on to show how transfer and simplification in-teract in our data.

During the many years of interest in the phenomenon of language transfer,one of the most frequently discussed issues has been transferability. Againand again researchers have addressed the question of what constraints, if any,there are on cross-linguistic influence. Some have stated or implied that virtu-ally no constraints exist (e.g., Thomason & Kaufman, 1988), whereas others

An earlier version of this paper was presented at the Second Language Research Forum, October,1998. The authors would like to thank Kathleen Bardovi-Harlig, Donna Lardiere, Jussi Niemi, BonnieSchwartz, Helena Sulkala, Sangeetha Venkatraman, and Don Winford for their help with different as-pects of this research.

Address correspondence to: Scott Jarvis, Dept. of Linguistics, Ohio University, Gordy Hall 383,Athens, OH 45701; email: [email protected].

2000 Cambridge University Press 0272-2631/00 $9.50 535

Page 2: Jarvis & Odlin 2000

536 Scott Jarvis and Terence Odlin

have seen the constraints as very far reaching: In the case of Dulay, Burt, andKrashen (1982), for example, transfer involving either morphology or syntaxwas considered to be minimal. Although many problems remain to be solved,the growth of the empirical record has made it increasingly clear that somesupposedly nontransferable structures do indeed find their way into interlan-guage grammars: for example, basic word-order patterns and semanticallyopaque idioms (e.g., Odlin, 1989, 1990, 1991). Our paper addresses yet anothersupposedly nontransferable domain: bound morphology in the context of amore general investigation of transfer in spatial reference.

The term “morphological transfer” can mean many things, and in the nextsection these possibilities are considered as well as the notion of transfer it-self. Of all the subsystems of language, there has probably been more skepti-cism about the transferability of morphology than about any other. Writing in1881, William Dwight Whitney downplayed the likelihood of such “transfer”(his term), especially transfer involving inflectional morphology, and argu-ments similar to his have found their way into the claims of other studentsof historical linguistics and language contact, several of whom are noted byThomason and Kaufman (1988). During the 1970s, when many second lan-guage researchers felt skeptical about grammatical transfer, morphology wassometimes pointed to as an area immune to cross-linguistic influence. For Du-lay and Burt (1974), the investigations of Uriel Weinreich (1953/1968) came upwith no convincing evidence for morphological transfer. Krashen (1983) madesimilar claims, and in recent papers Eubank (1993/1994; Eubank, Bischof, Huff-stutler, Leek, & West, 1997) has adopted the same position. In Eubank’s case,the supposed facts about nontransferability underpin certain positions hetakes with respect to Universal Grammar. In his 1993/1994 paper, he arguedthat “NL influence appears to be more limited: Lexical as well as functionalprojections transfer, as do directionality characteristics of those projections,but morphology-driven information like the strength of the inflection does nottransfer” (pp. 183–184). Further on in the same paper, he cited previous work:“The Dulay and Burt research lends support to the idea that inflection doesnot transfer even though other aspects of the NL may” (p. 206). In his co-au-thored 1997 paper, some of the arguments about UG had changed, but theposition on nontransferability remained the same, with an assumption that“overt inflectional morphology generally does not transfer from NL to L2”(p. 176).1

Our differences with Eubank have nothing to do with the UG issues he hasraised but rather with the empirical issue of transferability itself. In general,three points about the tradition of skepticism about this issue should be em-phasized. First, any discussion of morphological transfer should consider allof the possible ways in which cross-linguistic influence might be manifested—and this is something that the skeptics have generally failed to do. Second, acareful reading of Weinreich (1953/1968, pp. 33, 68) shows that he was not soskeptical about such transfer; in fact, some of the evidence he cited indicatesthat morphology is susceptible to cross-linguistic influence (e.g., Rosetti,

Page 3: Jarvis & Odlin 2000

Morphological Transfer 537

1945–1949). Third, since Weinreich’s time there have been studies that makea convincing case for transferability, some of which will be discussed in thefollowing sections.

The empirical issue at present is not just whether morphological transferoccurs but also what constitutes this type of transfer and how such transfermay operate. As a contribution toward understanding those questions, our in-vestigation indicates that the nominal case system of Finnish is a frequentsource of cross-linguistic influence in the acquisition of English. The evidenceseems especially compelling from a more general comparison of the systemsof spatial reference used by learners whose native language (L1) is Swedishand by those whose L1 is Finnish. As will be seen, the patterns of spatial refer-ence favored by these two populations are quite different and clearly reflectdifferences in the L1s. Before the comparison, however, we discuss the possi-ble meanings of morphological transfer, which is followed by a comparison ofthe morphosyntax of spatial reference in Finnish and Swedish. We then de-scribe the investigation and the methods used, with the discussion of the re-sults offering a chance to consider closely issues of transfer and simplificationin relation to earlier work.

POSSIBLE MEANINGS OF MORPHOLOGICAL TRANSFER

As Selinker (1992) observed, the term “language transfer” encompasses a widerange of phenomena. Yet although its status as a cover term is undeniable,much of what is called cross-linguistic influence can be viewed in terms ofretentions: Whenever challenges of using or understanding a second languagearise, learners may retain something from their L1 or some other language toaid in coping with the new challenges. The notion of retention is applicablewhether or not the attempt at coping converges with the target language (i.e.,positive transfer) or diverges (negative transfer). Although the notion of reten-tion might seem equivalent to claims by Krashen (1983) that transfer is nomore than “falling back” on the L1, Krashen’s claims are not credible eventhough some transfer might be regarded as a default behavior arising fromignorance of the target language (L2). One key way in which retention differsfrom falling back is that learners often create hybrid forms that show a blend-ing of L1 and L2 forms, as seen in a sentence written by a native speaker ofSwedish who was describing a film: They have many appletrees and vinetrees.The use of vinetrees owes something to the Swedish word vintrad “grape vine,”literally “wine tree,” but it is clearly a blend of an L1 and an L2 form. Suchlexical hybrids reflect only one of the ways in which any learner can creativelyconstruct an interlanguage shaped by many factors, including the L1 (seeSchachter, 1983, 1992; Selinker, 1992; Sharwood Smith, 1979). Although reten-tion encompasses much of what is meant by language transfer, it is not neces-sarily applicable to all interlanguage behaviors that may be susceptible toinfluence from the L1, including avoidance, hypercorrection, and simplifica-

Page 4: Jarvis & Odlin 2000

538 Scott Jarvis and Terence Odlin

tion, the latter behavior being one that we will discuss in some detail later inthis article.

Even within the restricted scope of retention, however, the notion of mor-phological transfer includes many phenomena. For example, cross-linguisticinfluence involving any free or bound morpheme technically qualifies as mor-phological transfer. On the other hand, rather different issues can be involvedwhen bound morphemes are in question as opposed to independent words.Transfer involving bound morphology can itself be subcategorized in severalways. Second language users may make interlingual identifications that in-volve both semantic and phonological structures. For example, an Estonianlearner of Finnish might identify, quite correctly, the Finnish inessive case in-flection -ssa with the Estonian form -s and thus equate Finnish taivaassa “inheaven” with Estonian taevas “in heaven,” with positive transfer resulting inspeech or writing if the extra syllable in the Finnish form is discarded.

Interlingual identifications involving bound morphology may also be possi-ble even when the L2 structure does not rely on affixes. On a purely semanticbasis, a Finn might equate taivaassa with English in heaven, where the freemorpheme in functions as the equivalent of the inessive case marker in Finn-ish. Whereas this example involves positive transfer, negative transfer mightalso occur—and indeed, examples of such divergence are discussed at somelength later in the paper.

Bound morphology itself is not at all monolithic. The distinction betweeninflectional and derivational morphology is clearly important, and Whitney(1881, p. 20) considered inflectional morphemes to be less likely candidatesfor transfer than derivational ones, which he also considered to be more orless immune to transfer. Since Whitney’s time, however, a number of counter-examples to his claim have come to light and, once again, some of the evi-dence to be discussed later (e.g., Sulkala, 1996) provides still furtherinstances.

For research on L2 acquisition, the focus of transfer research is naturallyon retentions from the L1 (or some other previously learned language). How-ever, cross-linguistic influence can also work in the opposite direction, wherethe acquisition of L2 structures causes some kind of change in the L1. Amongthe kinds of transferable structures is bound morphology. One of the mostconvincing counter examples to Whitney’s (1881) claim about the nontransfer-ability of inflections is a study by Dawkins (1916) of dialects of Greek in whichthere are clear instances of Turkish verbal inflections finding their way intothe verb systems of Greek, such influence resulting from the bilingualism ofGreeks who had acquired Turkish. Although such cases are only indirectly rel-evant to SLA research, they should inform any well-developed characteriza-tion of L1 transfer because in both “borrowing” and “substratum” transfer (touse terms from Odlin [1989], with the latter covering cases of transfer in SLA),there must be interlingual identifications, and the “latent psychological struc-ture” underlying identifications in either type of transfer may be largely thesame (see Selinker, 1992).

Page 5: Jarvis & Odlin 2000

Morphological Transfer 539

As suggested already, the outcome of any cross-linguistic influence can beeither convergent or divergent, and it is important to emphasize that anyclaims about the transferability of bound morphology must consider positiveas well as negative transfer. Eubank’s (1993/1994) claim about the “overtbound morphology” is clearly untenable in light of a study by Orr (1987) thatshows that speakers of Ngoni, a Bantu language, had a strong advantage inlearning the complex prefixation of Chichewa, another Bantu language, in com-parison with learners of Chichewa whose L1 was Gujarati, an Indo-Europeanlanguage that makes no use of the Bantu prefixing system. Such comparisonsof different L1 groups can help to determine negative transfer, but they areespecially valuable in demonstrating a facilitative influence as in the case ofthe advantage shown by Ngoni speakers (see Odlin, 1989). Because such com-parative studies are the exception and not the rule in language-transfer re-search, it is not surprising that possible facilitating effects of similarities ofbound morphology in some cases have gone largely unnoticed.

Just as positive transfer has often been ignored in discussions of morpho-logical transfer, so too have hypotheses concerning comprehension (in con-trast to production) been rarely discussed. That is, any fully satisfactoryaccount of transfer and morphology should specify just how cross-linguisticsimilarity might contribute advantages in listening or reading comprehension.Intuition suggests, for example, that the similarity between Finnish taivaassaand Estonian taevas would help an Estonian in comprehending Finnish just asit presumably would in producing structures in the L2. More generally, if Ring-bom (1992) is correct that the advantages of transfer are especially great incomprehension tasks, the benefits may frequently involve some kind of posi-tive morphological transfer.

Whatever effects transfer may show with regard to a morphological sys-tem, there are also possible effects arising from the interaction between otherlinguistic subsystems. For example, the extensive inflection on Finnish nouns(see the section titled “Typological Comparison of Finnish and Swedish”) in-teracts with the order of clause constituents so that the word order of Finnishis highly flexible. Thus when Finnish learners produce word-order errors inEnglish, as in She took from the car bread, the interdependence of morphologyand syntax in Finnish may well be a factor. To our knowledge, such kinds ofinteractive factors have not yet been extensively studied with regard to trans-fer, and they are only briefly mentioned here.

PURPOSES OF THE PRESENT INVESTIGATION

The present paper examines morphological transfer in the spatial referenceof Finnish-speaking and Swedish-speaking learners of English. Our purpose isthreefold: (a) to examine the types of interlingual identifications that learnersmake between the spatial morphology of the L1 and L2, (b) to investigate theformal means by which learners exhibit their interlingual identifications, and(c) to explore possible interactions between morphological transfer and mor-

Page 6: Jarvis & Odlin 2000

540 Scott Jarvis and Terence Odlin

phological simplification in learners’ spatial reference. Our research is guidedby the following questions:

1. Can learners from different L1 backgrounds be distinguished according to thetypes of options that they pursue with respect to both spatial reference and mor-phological transfer?

2. Will Finnish-speaking learners make interlingual identifications between the boundlocative morphology of Finnish and the spatial prepositions of English?

3. What types of morphological transfer will Finnish-speaking learners exhibit in theirEnglish spatial reference, and how will they differ from Swedish-speaking learners?

4. Will both Finnish-speaking and Swedish-speaking learners exhibit morphologicalsimplification in their English spatial reference, and will they differ with respect totheir patterns of simplification?

In cases in which learners retain structural or semantic properties fromtheir L1s, it follows that learners from different L1 backgrounds will have attheir disposal different (though probably overlapping) sets of structural andsemantic options for expressing their intended message in the L2. Thus, animportant question from the perspective of the present paper is whether thistype of transfer can be found in learners’ reference to spatial relations, andwhether it can involve the retention of L1 morphological properties. As men-tioned earlier, the present study compares the English spatial expressions ofFinnish-speaking learners with those of Swedish-speaking learners. Becausethe spatial systems of the learners’ L1s differ both structurally and semanti-cally (see section entitled “Typological Comparison of Finnish and Swedish”),it is relevant to ask whether these differences will lead to differences in thestructural and semantic patterns (or options) they display in their Englishspatial reference. This is essentially what the first research question asks.

The second question more explicitly addresses the transferability of boundmorphology. This question becomes pertinent because, although Swedish issimilar to English in its coding of locative and directional information primar-ily through spatial prepositions, Finnish expresses spatial relationships mainlythrough bound, agglutinative morphology. If one were to adopt the traditionalassumption that bound morphology does not transfer, then one might expectto find transfer in Swedes’ but not Finns’ English spatial reference. This alonecould result in different sets of options for Finnish-speaking versus Swedish-speaking learners, which would constitute positive support for our first re-search question. Our second research question, however, requires us to ana-lyze more specifically whether properties of Finnish bound morphology areactually retained in Finns’ reference to spatial relations in English.

The third research question requires a full account of the structural andsemantic patterns that both Finns and Swedes exhibit in their English spatialreference. It also requires a comparison of the options that Finns and Swedesseem to have at their disposal. Finally, this question requires an interpretationof which options involve transfer, and, moreover, which ones involve morpho-logical transfer.

Page 7: Jarvis & Odlin 2000

Morphological Transfer 541

The final research question addresses the interaction between simplifica-tion and transfer. First, it asks whether learners’ patterns of spatial referenceinclude instances of structural simplification (e.g., omissions) or semantic sim-plification (e.g., semantic overgeneralization). It also asks whether learnersfrom different L1 backgrounds differ in the patterns of simplification that theyexhibit. Insofar as performance differences between learners from different L1backgrounds suggest the possibility of transfer, any emerging differences inFinns’ and Swedes’ patterns of simplification may indicate an interaction be-tween transfer and simplification. In other words, it may indicate that sometypes of simplification come about at least partly through transfer. We con-sider these possibilities in more detail in the “Discussion” section.

TYPOLOGICAL COMPARISON OF FINNISH AND SWEDISH

To fully appreciate the aims and results of the present study, one must firstunderstand how our learners’ L1s differ. In this section, therefore, we providea brief typological comparison of Finnish and Swedish. As a point of depar-ture, it is important to recognize that the Finnish and Swedish languages arehistorically unrelated and typologically distinct. Finnish is a Finno-Ugric lan-guage, closely related to Estonian, and more distantly to (among other lan-guages) Hungarian. Swedish, however, is a Germanic language and is closelyrelated to (among other languages) Danish, Dutch, and English. AlthoughFinns and Swedes share a long history of cultural contact, the structural char-acteristics of their languages have remained largely unaffected. Some borrow-ing, mainly lexical, has occurred—especially from Swedish to Finnish—butthis has had little effect on the grammar of either language (cf. Ringbom, 1987,pp. 19–23).

One of the most striking grammatical characteristics of Finnish is its nomi-nal case system (see Homburg & Nikanne, 1993; Karlsson, 1983). Finnish has15 productive nominal cases that are expressed as agglutinative suffixes onnouns and their modifying adjectives. (It should be noted, however, that theagglutination in Finnish sometimes involves obscure stem-affix boundaries,unlike Turkish; see Comrie, 1981.) Most of the cases are expressed differentlyin the singular and plural, with the result that almost every noun and adjectivein Finnish has over two dozen distinct forms—and even more when the Finn-ish possessive suffixes and agglutinative particles are taken into consider-ation. The nominal cases in Finnish include the nominative, accusative,partitive, genitive, and a number of cases that correspond roughly with vari-ous prepositions and other function words in either English or Swedish (see,e.g., Karlsson, 1983, pp. 22–24).

Swedish, like English, has only the nominative and genitive cases for nouns,and the nominative, accusative, and genitive cases for pronouns. The Swedishverbal system is also simple and has no subject-verb agreement at all (e.g.,Beite, Englund, Higelin, & Hildeman, 1963). Finnish, on the other hand, has acomplex subject-verb agreement system in which the verb takes a separate

Page 8: Jarvis & Odlin 2000

542 Scott Jarvis and Terence Odlin

Table 1. Basic spatial systems of Finnish, English, and Swedish

English and SwedishFinnish locative cases basic spatial prepositions

Goal/Locatives Goal Location Source Goal location Source

Internal -Vn -ssa -sta into in out (of)in i i (ut) ur/ifran

External -lle -lla -lta onto on off (of)pa pa av

Neutral to at fromtill vid fran

suffix depending on whether the subject is first, second, or third person andwhether it is singular or plural (e.g., Karlsson, 1983).

Swedish is grammatically richer than Finnish only in its grammatical gen-der, article, and prepositional systems. Swedish has two grammatical genders,whereas Finnish has none. Swedish also has both definite and indefinite arti-cle systems that realize the same functions as English articles and additionallyreflect the grammatical gender and number of the noun (Beite et al., 1963).The article systems of Swedish will not be discussed further in this paper, butit is relevant to point out here that the definite article in Swedish involves abound morpheme. Thus Swedish has its own distinct complications involvingbound morphology. Finnish, although optionally allowing for the use of third-person pronouns as definite articles in certain contexts, lacks articles as aseparate grammatical class (Ringbom, 1987; cf. Laury, 1997). Finally, whereasSwedish has a rich system of prepositions to represent spatial, temporal, andother grammatical and semantic relations, Finnish has only about 15 preposi-tions, most of which are more commonly used as postpositions. Most of thespatial and temporal relations that are expressed in Swedish and English withprepositions are expressed in Finnish through the nominal case system, andnearly all others are expressed with postpositions. The locative case systemin Finnish is shown in Table 1, and is contrasted with the basic locative prepo-sitions in Swedish and English.

The six Finnish locative cases are formally and semantically divided intotwo general classes: the internal and the external locatives. Each class is fur-ther divided according to whether the marked noun represents the goal, loca-tion, or source of the predicated action. In English and Swedish, prepositionscan be used to represent these spatial distinctions, and there are at leastthree important differences with respect to the spatial systems of English andSwedish versus Finnish. First, unlike Finnish, English and Swedish include aclass of prepositions (e.g., to, at, from) that are neutral as to whether the goal,location, or source is internal or external. Second, unlike Finnish, English andSwedish conflate goal and location into a single category for internal (e.g., in),

Page 9: Jarvis & Odlin 2000

Morphological Transfer 543

external (e.g., on), and neutral (e.g., at) prepositions. Third, whereas the goaland location suffixes in Finnish are formally similar to their source counter-parts (e.g., the formal similarity between the external suffixes -lle, -lla, and-lta), the goal and location prepositions in English and Swedish are formallydistinct from their source counterparts. Examples of these contrasts will ap-pear later in the tables of our empirical study.

METHOD AND ANALYSIS

Participants in the present study included 140 Finnish and 70 Finland-Swedishadolescents who were tested in English, as well as another 66 Finnish, 44 Fin-land-Swedish, and 66 American adolescents who were tested in their L1s (seeJarvis, 1998).2 The Finns were fifth, seventh, and ninth graders in publicschools in Finland where the language of instruction is Finnish, and the Fin-land Swedes (henceforth, Swedes) were seventh and ninth graders in similarschools in Finland where the language of instruction is Swedish (see Ringbom,1987). The Americans were in the fifth, seventh, and ninth grades. All of theEFL learners were in their third, fifth, or seventh year of formal English in-struction at the time of testing. As part of a larger study, participants wereshown an 8-minute segment of the silent film Modern Times, and immediatelyafterwards they were asked to write a narrative of the film. These narrativesconstitute the data for the present paper.

Two types of analysis were performed. First, we identified the directional,goal-oriented usages of the verbs take, sit, and put. While doing so, we tabu-lated participants’ use or nonuse of prepositions along with these verbs, andwe quantified the results according to lexical type and participants’ L1 back-ground. In the second analysis, we identified participants’ reference to a spe-cific scene in the film in which a person is seen taking bread out of the backof a delivery vehicle. In connection with this analysis, we tabulated and quan-tified learners’ reference to this source-oriented spatial relationship. The re-sults of the two analyses are presented in the next section.

RESULTS

We will first discuss the two verbs that we have examined in detail: take andsit (although there will also be a brief discussion of put). Both of these occurfrequently in learners’ descriptions, quite often with locative and directionalexpressions. We coded each occurrence of any recognizable form of sit—notonly sit, sits, sat, and sitting, but also forms such as sett and satted. Likewise, inthe case of take, we coded each occurrence of any recognizable form, includ-ing take, takes, took, taking, and taken, and also variant forms such as toke andtoken.

In the case of sit, there were 352 recognizable forms, and of these 239 hadlocative or directional phrases that included nouns or pronouns. We excludedfrom the analysis any cases of sit that did not occur with an internal nominal

Page 10: Jarvis & Odlin 2000

544 Scott Jarvis and Terence Odlin

Table 2. Examples of prepositions co-occurring with sit

Preposition Sample sentence

Zero C.C. and the woman go to sit the grass.in After that they sat in grass.on Charlie and woman sit on the crass

[sic].under Then they sat under the big tree.by Chaplin and the woman sit by the street.to Chaplin and girl sat to grass.

Table 3. Frequency of prepositions co-occurring with sit

Group Zero in on under by to Other

Finns 12 14 43 18 2 8 15Swedes 0 17 18 21 0 1 15Americans 0 7 15 12 15 0 5Total 12 38 76 51 17 9 35

Table 4. Frequency of prepositionsco-occurring with sit and grass in thelawn scene

Group Zero in on to

Finns 5 5 21 6Swedes 0 7 5 0Americans 0 4 1 0Total 5 16 27 6

argument, which were generally either simple intransitive verbs or tokens oc-curring with an adverb such as there or down. However, we did include allcases in which the adverb co-occurred with a prepositional phrase as in sitdown on the grass. Table 2 shows sentences with the most common examplesof prepositional choices made by the native speaker and learner groups, andTable 3 shows the statistical pattern.

The figures in Table 3 show some striking intergroup differences. Most sig-nificantly, only the Finns used patterns with zero prepositions. The 12 in-stances of zero came from 12 different individuals, and these omissions ofprepositions reflect an important intergroup difference. Likewise, Finns moreoften chose the preposition to, whereas their choice of under does not seemvery distinctive. In the case of in and on, Table 3 suggests possible intergroupdifferences, but these can be more clearly appreciated when only instancesinvolving the same context are considered. Thus, Table 4 shows the choices

Page 11: Jarvis & Odlin 2000

Morphological Transfer 545

Table 5. Examples of prepositions co-occurring with take

Preposition Sample sentence

Zero The policeman take Charlie policecar.to Police take him to a police car.into Then he take Charlie into the car.in Then policeman take Chaplin in the police car.

Table 6. Frequency of prepositions co-occurringwith take

Group Zero to into in Other

Finns 10 25 10 11 1Swedes 0 13 2 6 0Americans 0 7 1 0 1Total 10 45 13 17 2

when the young writers attempted to recount a scene in the film where thetwo escapees sit on a lawn and dream of a better life. The most frequent nounused in nominal arguments in this context was grass, as in the first three exam-ples in Table 2. From Table 4 it is evident that the Finns showed a strong pref-erence for phrases such as on the grass, whereas the Swedes preferred in thegrass. A chi-square test run on the Finns’ and Swedes’ choices revealed a sta-tistically significant intergroup difference (χ2 = 11.39, df = 3, asymptotic p =0.01). To control for the possible effects of small cell values, a Fisher’s Exacttest was also run (exact p = 0.01), which confirmed the results of the chi-square test.

In the case of take, there were many more recognizable forms, 771 in all.However, we excluded numerous cases from our analysis. Take is usually atransitive verb, yet although it has an obligatory direct-object argument, loca-tive arguments are normally optional. In our corpus, this verb showed propor-tionally fewer instances with a nominal locative argument, in comparison withsit. Moreover, we excluded nominal arguments involving source—for example,take milk from a cow. These instances were interesting, and we will discusslearner expressions involving source later; however, our count focuses oncases where take occurred with a nominal argument indicating a spatial goal,as in took Charlie Chaplin to the policecar. Table 5 shows sentences with themost common examples of prepositional choices, and Table 6 shows the sta-tistical pattern.

As in the case of sit, there are numerous instances of zero prepositions oc-curring with take produced by the Finns: 10 cases, with 9 of them coming fromdifferent individuals. Moreover, there is a strong preference among the Finnsfor into, which was used 10 times, whereas Swedes only used that preposition

Page 12: Jarvis & Odlin 2000

546 Scott Jarvis and Terence Odlin

Table 7. Frequency of prepositionsco-occurring with take and -car in thepaddy wagon scene

Group Zero to into in

Finns 8 16 7 8Swedes 0 9 1 5Americans 0 1 0 0Total 8 26 8 13

twice. Though both groups used to numerous times, the choice of in is alsocommon, as in Then policeman take Chaplin in the police car. We will discussthe unique status of in a little further on. The majority of nominal goals co-occurring with take refer to the scene where Chaplin is taken to a paddywagon, and so the statistical pattern in Table 7 is similar to what is seen inTable 6. Even though a chi-square test and a Fisher’s Exact test failed to reveala statistically significant difference between the Finns and Swedes with re-spect to the figures in Table 7 (χ2 = 5.59, df = 3, asymptotic p = 0.13, exact p =0.16), one can nevertheless see from the results that only the Finns exhibit thezero-preposition option in this context.

We suspect that positive transfer from both Finnish and Swedish accountsfor the less divergent patterns in the two L1 groups with take in comparisonwith sit. Although eight Finns used zero prepositions, the three choices of to,into, and in are all viable in the L2, and the structure of the L1 can motivateall three choices (though in the case of in, the motivation is somewhat morecomplex, as discussed in the Discussion section). In the L1 Finnish composi-tions written by students in the control group, the illative case (representinginternal goal) is the usual inflection to mark the paddy wagon (poliisiauto) asa directional referent, as in Poliisi vei Chaplinin poliisiautoon “The policemantook Chaplin to the police car,” where the -on inflection marks the illativecase. The normal translations of illatives are to and into, and so it is not sur-prising that most Finns would be able to make successful interlingual identifi-cations. The likelihood of L1 influence is all the greater in the case of one Finnwho apparently could not recall the English form took and used the Finnishequivalent (vei) instead: Policeman came and vei the Chaplin to policecar. Inthe L1 Finnish compositions, vei is used several times for the paddy wagonscene, and every time it is so used, the Finnish word for police car is markedwith the illative (e.g., poliisiautoon). In the case of the Swedes, their L1 hasprepositions corresponding fairly closely to the two most common Englishchoices till for English to, and i for English in (see Table 1), and these choicesare also evident in the L1 Swedish compositions.

We focused on take and sit because both of these verbs frequently co-occurwith locative nominals and because they show significant intergroup differ-ences in the choice of prepositions (including zero). We should note, how-

Page 13: Jarvis & Odlin 2000

Morphological Transfer 547

Table 8. Examples of source prepositional choices co-occurringwith bread in the theft scene

Preposition Sample sentence

Zero The girl stole a loaf a bread the car and run away . . .in She stealing in the car a bread.out of So when he goes in she takes some bread out of the truck

and runs.from Then she take a bread from car and run away.

Table 9. Frequency of source-oriented choices co-occurring with bread inthe theft scene

Internal External Neutral

Goal Loc. Source Goal Loc. Source Goal Loc. SourceGroup Zero into in out of onto on off to at from

Experimentalgroups

Finns 2 1 6 2 1 1 0 0 2 12Swedes 0 0 1 1 0 1 0 0 0 13Americans 0 0 0 5 0 0 1 0 0 4

L1 controlgroups

Finns 0 0 0 26 0 0 0 — — —Swedes 0 0 0 6 0 0 0 0 0 9

ever, that intergroup differences are also evident in the case of other verbs aswell. With regard to the verb put, for example, the preposition to was used byfive Finns (and by no Swedes), as in the sentence . . . a policeman put him tothe policecar. The corpus indeed shows intergroup differences in many areasbesides the coding of spatial concepts.

Our second analysis is an examination of learners’ spatial reference to ascene in the film in which a person is seen taking bread out of the back of abakery delivery truck. There were 53 identifiable references to this event inthe English narratives, 26 in the Finnish narratives, and 15 in the Swedish nar-ratives. We coded each occurrence of participants’ reference to this event ac-cording to the prepositions or locative markers they chose to refer to thesource relationship between the bread and the delivery vehicle. Table 8shows sentences that are representative of the most common prepositionalchoices made by the native speaker and learner groups, and Table 9 showsthe statistical patterns of the native speaker and learner groups, as well as ofthe corresponding L1 control groups.

The figures for the Finnish learner group show some striking differences

Page 14: Jarvis & Odlin 2000

548 Scott Jarvis and Terence Odlin

Table 10. Examples of Finns’ nonstandard use of in categorized according tothe Finnish spatial system

Finnish case suffix Sample sentence produced by L1-Finnish learner

-Vn (internal goal) She runaway, but bumped in Chaplin and they both fell down.-ssa (internal location) They lay in the crown [ground], and get up.-sta (internal source) When they had escaped in the police car they sat under the tree.-lle (external goal) That woman who stole bread Charlie give he place in woman.

both from the Swedish and American groups, on the one hand, and from theFinnish control group, on the other. As before, the Finnish learner group isthe only group to use patterns with zero prepositions. The native speaker andL1 control groups are the most consistent in choosing only prepositions orlocative suffixes that conventionally mark source relations, whereas thelearner groups—most notably the Finns—use a wider variety of prepositions.The most frequent choice for both the Finnish and Swedish learner groups isfrom. The 12 instances of from for the Finnish group came from 11 differentindividuals, and the 13 instances of from for the Swedish group came from 13different individuals. What is most striking about these figures is that a strongsecond choice for the Finns is the preposition in. This is striking because indoes not seem to be a logical candidate to represent the source relation thatwas ostensibly being referred to, and it is also striking because it does notseem to be motivated by the internal-source locative suffix that was consis-tently chosen by the Finnish L1 control participants.

To further investigate Finns’ unexpected use of in, we conducted a follow-up analysis of the data in which we identified all of the nonstandard uses ofin by the Finnish learner group. Of the 364 instances of this preposition in theFinns’ narratives, we identified 64 nonstandard occurrences. Seven of theseinvolved the phrase in there, and five others were conspicuous for syntacticreasons. The remaining 52 nonstandard occurrences—produced by 35 differ-ent individuals—involved a variety of nontargetlike uses of in for example,Charlie Chaplin is in her way and they crash to each others and fall in the groundand It was some meal and they were cutting it in same time (emphases added).In an attempt to make sense of these, we categorized them according to theFinnish locative cases that were used in the same spatial reference contextsby the Finnish L1 control group. The results are revealing. Table 10 showsexamples of sentences categorized in this manner, and Table 11 shows thestatistical pattern. The 21 instances of in categorized as internal goal wereproduced by 19 different individuals, the 8 instances in the internal locationcolumn were produced by 6 different individuals, the 21 instances of in as aninternal source were produced by 19 different individuals, and the 2 instancesof in as an external goal marker were produced by two different Finnish learn-ers of English.

It is important to point out that the data elicited from the Finnish learners

Page 15: Jarvis & Odlin 2000

Morphological Transfer 549

Table 11. Frequency of Finns’ nonstandard use of in by comparisonwith the Finnish spatial system

Internal External

Goal Loc. Source Goal Loc. Source

Case suffix -Vn -ssa -sta -lle -lla -ltaFinns (use of in) 21 8 21 2 0 0

contain many more instances of in used as an internal goal or internal locationmarker than are shown here. These figures represent only the Finnish learn-ers’ nonstandard uses of the preposition in. Given that this preposition carriesthe meaning of both internal goal and internal location in Standard English,the figures in the first two columns of Table 11 represent instances in whichFinnish learners have overgeneralized the use of in to contexts that are notconventionally marked as internal goal or internal location by native Englishspeakers. The most striking result, of course, is the figure in the third columnof Table 11 that reflects the large number of Finnish learners who used in tomark an internal source. This is conspicuous because, on the surface, thelearners’ overgeneralization of in in this way does not seem likely to be moti-vated by either direct interlingual identification or L2 input. Nevertheless, it isnoteworthy that Finns’ nonstandard use of in is limited almost exclusively tothe internal spatial relations. The evidence here seems to suggest that oneway in which Finnish learners of English overgeneralize English spatial prepo-sitions is by choosing a single preposition, in in this case, to represent an en-tire class of spatial relations, here the internal spatial system. Thisassumption is given further support by the fact that each of the 19 Finns whoused in to represent an internal source relation also used it to mark an inter-nal goal or internal location relation, or both. We will say more about theFinns’ overgeneralization of in in our discussion of the results, to which wenow turn.

DISCUSSION

Research Questions Addressed

Three of our four research questions receive direct confirmation from thepresent results: (a) Finns and Swedes do show differences in the options theypursue with respect to both spatial reference and morphological transfer, (b)Finns do seem to make interlingual identifications between the bound locativemorphology of Finnish and the spatial prepositions of English, and (c) Finnsand Swedes do differ noticeably in their patterns of morphological simplifica-tion. The remaining research question is more exploratory, restated here asfollows: What types of morphological transfer do Finnish-speaking learners ex-

Page 16: Jarvis & Odlin 2000

550 Scott Jarvis and Terence Odlin

hibit in their English spatial reference, and how do they differ from Swedish-speaking learners? We attempt to answer this question—as well as providemore support for the previously mentioned research questions—in the follow-ing paragraphs.

To begin, the striking differences we have found between the Finns andSwedes in their reference to locative and directional expressions in Englishleave little doubt that L1 influence is at play in this area of second languageacquisition. The differences arising from L1 background concern the prefer-ence for different prepositions by the Finns and Swedes in the same spatialcontexts, as well as the omission of spatial prepositions by the Finns but notthe Swedes. One of the clearest examples of differing prepositional choicesinvolves the learners’ reference to the scene where Chaplin and his femalefriend are sitting on the grass. The Finns show a strong preference for on,whereas the Swedes prefer in (although both choices are in fact grammatical).Corresponding patterns were found for the L1 control groups: 27 instances of-lle/-lla “on” and 0 instances of -ssa/-Vn “in” by the Finns, and 11 instances of i“in” and 5 instances of pa “on” by the Swedes. This is an important finding fortwo reasons. First, it suggests a strong role for semantic transfer in learners’spatial reference and, second, it shows that Finns, in particular, are capableof making interlingual identifications between postposed bound morphologyin Finnish and preposed free morphology in English. The basis for these inter-lingual identifications is semantic—that is, learners match meanings in the L1and L2, ignoring the formal disparities.

With respect to the omission of prepositions, the Finnish learner groupused patterns with zero prepositions in all of the spatial contexts that we ex-amined, whereas the Swedish learner group did not use a zero preposition inany of these contexts. Although the use of a zero preposition is a form of lin-guistic simplification, its use by the Finns also constitutes a form of transfer,given that the structural nature of the Finnish locative cases predisposesFinns to disregard preposed function words as relevant spatial markers. Thus,the Finns’ omission of spatial prepositions in English seems to arise out of aninteraction between simplification and transfer.

Another result from our analysis that appears to have arisen out of the in-teraction between transfer and simplification is the Finns’ overgeneralizationof the preposition in. Overgeneralization, like omission, is a form of simplifica-tion—as we discuss later—but the way the Finns overgeneralize in as a coverterm for internal locative and directional relations seems to be motivated bystructural differences between Finnish and English. As mentioned earlier, theFinnish locative cases distinguish between external and internal relations, aswell as between goal, location, and source relations. English makes similar dis-tinctions with its spatial prepositions, but the English prepositions that ex-press location (in, on, and at) are also used to express goal relations. Thismeans that the preposition in in Standard English conflates both internal loca-tion and internal goal. The Finnish locative cases, in contrast, do not allowsuch semantic conflation. Therefore, when faced with L2 input where the

Page 17: Jarvis & Odlin 2000

Morphological Transfer 551

preposition in is used to express more than one internal relation, some Finn-ish-speaking learners of English may assume that in can be used to express allinternal relations. An extension of Andersen’s (1984) One-to-One Principle maybe applicable here: If a linguistic form does not represent a single concept,such as internal location, then it must represent a single class of related con-cepts, such as internal space. The Swedes do not overgeneralize in in this waypresumably because the corresponding Swedish word, i, conflates internal lo-cation and internal goal in a similar manner to English.

Relevance to Previous Research

The conclusions just discussed are consonant with previous work on the ac-quisition of spatial reference. Because prepositions are a notoriously difficultarea of English for nonnative speakers, there have naturally been several in-vestigations (e.g., Carroll, 1997; Mukattash, 1984; Pavesi, 1987; Ringbom, 1987;Schumann, 1986). Not surprisingly, these studies have also identified cases oftransfer. Mukattash, for example, found two sources of negative transfer in theresults of a standardized test taken by Jordanian learners of English: Jorda-nian Arabic and Modern Standard Arabic. Although many of his learners’ prep-ositional errors are not directly attributable to transfer, Mukattash assertedthat they are nevertheless “dominated by L1 interference” (p. 59) and largelyconcern the use of primary counterparts (i.e., perceived lexical translationequivalents). Pavesi likewise saw transfer as one, although not the sole,source of some of the English prepositional errors of Italian speakers, but sheprovided few details on the exact nature of the cross-linguistic influence. In-vestigations into the use of English prepositions by Finnish-speaking andSwedish-speaking learners, in contrast, have been much more explicit aboutthe nature of cross-linguistic influence (see Ringbom). One particular point ofconcurrence between earlier research in Finland and our own is that Finnsomit English prepositions far more often than Swedes do, and this appears tobe largely due to the more substantial structural differences between Finnishand English with respect to the languages’ spatial systems.

Not all the work on transfer involving prepositions has focused on English,however. For example, Harley (1989) compared the performance of native andnonnative speakers of French in their use of spatial prepositions in a con-trolled composition task. The nonnative speakers showed a strong preferencefor using prepositions in a manner consistent with their English L1 and notwith the L2; for example, many writers chose the preposition a to express di-rectional concepts, as in Le chat courait a [sic] la maison “The cat ran to thehouse.” This use of a resembles the English use of to as a directional preposi-tion, and it differs from the French norm. In the target language, a is a locativebut not a directional, with jusqu’ a “to, till” being the normal directional as inLe chat a couru jusqu’ a la maison (the periphrastic verb a couru is also normal,although there was probably no interaction between the learner’s choice of theimperfect form courait and the choice of a). Although other semantic contexts

Page 18: Jarvis & Odlin 2000

552 Scott Jarvis and Terence Odlin

allow for positive transfer between English to and French a, the locative-direc-tional contrast elicited by Harley’s task led to negative transfer.

Along with work on the acquisition of prepositions, some recent researchhas examined problems that learners encounter in acquiring highly agglutina-tive languages such as Finnish (e.g., Martin, 1995; Sulkala, 1996). Sulkala’sstudy included a look at likely L1 influence from Estonian, another agglutina-tive language. The evidence for transfer identified by Sulkala clearly strength-ens our case about the transferability of bound morphology, but we shouldnote that in her study, as well as in Martin’s (which did not attempt to look atcross-linguistic influence), other factors besides transfer were also found tobe at work. By the same token, work on the acquisition of languages spon-sored by the European Science Foundation (e.g., Becker & Carroll, 1997; Ex-tra & Mittner, 1984) considered a variety of cognitive factors involved in theevolution of spatial reference in an L2, of which transfer was only one.

More on Simplification and Transfer

Our findings also support those of Schumann (1986), who found that transferand simplification together account for the structure of learners’ spatialexpressions. In particular, Schumann found that his one Cantonese- and twoJapanese-speaking learners, whose L1s lack spatial prepositions,3 producedstructures with zero prepositions much more frequently than his threeSpanish-speaking learners, whose L1 does contain spatial prepositions. Addi-tionally, Schumann’s Spanish speakers overgeneralized their use of the prepo-sition in to refer to spatial relations that would require on, at, or to in StandardEnglish. The Spanish speakers appear to have overgeneralized in on the modelof the Spanish preposition en, in which internal and external spatial meaningsare conflated. It is important to point out that Schumann’s Spanish speakersovergeneralized in in a substantially different way from how our Finnishspeakers did; these differences can be attributed to structural and semanticdifferences between the spatial systems of the learners’ L1s. In any case, ourfindings, along with those of Schumann, show that transfer and simplificationcan jointly contribute to the way learners refer to spatial relations in an L2.4

Like Schumann (1986), we identified two general types of simplification inlearner language that appear to interact with transfer. The first type is similarto what Meisel (1980) referred to as “restrictive simplification,” which he char-acterized as the use of structurally or grammatically reduced patterns (pp.36–37). In terms of spatial expressions, the clearest cases of restrictive simpli-fication would involve the omission of prepositions or related functional mor-phology. In our study, restrictive simplification in the form of prepositionalomissions was found to be frequent in the Finns’ spatial expressions but non-existent in the Swedes’. Our results, therefore, show not only instances of re-strictive simplification, but also a seemingly strong interaction between thistype of simplification and learners’ L1 background. Schumann and others (e.g.,Mukattash, 1984; Ringbom, 1987) have likewise found that even though restric-

Page 19: Jarvis & Odlin 2000

Morphological Transfer 553

tive simplification often occurs regardless of transfer, its occurrence is greatlyincreased when the two processes interact.

The second general type of simplification that we identified in our data issimilar to what Meisel (1980) referred to as “elaborative simplification” (cf. El-lis, 1982). Unlike restrictive simplification, elaborative simplification is notcharacterized by omissions or structural reduction but rather by an attemptat elaborating one’s interlanguage system. More precisely, elaborative simplifi-cation is an attempt “to formulate hypotheses about a certain rule which maybe approximations to the actual rule” (Meisel, 1980, p. 37). As elaborative sim-plification includes the notion of semantic overgeneralization, our resultsshow instances of this type of simplification, too. In particular, we demon-strated that Finnish-speaking learners often overgeneralize the use of in torepresent all internal spatial relations (goal, location, and source), whereasSwedish-speaking learners do not show the same tendency. Besides demon-strating that the two groups of learners differ with respect to their patterns ofovergeneralization, we have additionally shown that Finns’ hypotheses con-cerning the function of in appear to be directly motivated by the nature oftheir L1 spatial system. Thus, like Schumann (1986), we found convincingsupport for the interaction between transfer and overgeneralization, or elabo-rative simplification (for additional evidence of transfer in learners’ preposi-tional overgeneralizations, see, e.g., Correa-Beningfield, 1990; Ijaz, 1986). Insum, it appears that both restrictive and elaborative simplification can inter-act with transfer, although clearly not all instances of either restrictive orelaborative simplification involve transfer (see, e.g., Mukattash, 1984; Schu-mann, 1986), and not all instances of transfer involve simplification.

Although we adopted Meisel’s (1980) distinction, we certainly do not sharehis skepticism (1983) about the role of language transfer. Moreover, we find itimportant to emphasize that cross-linguistic influence is compatible with thatof hypothesis formation in relation to elaborative simplification. Indeed,Schachter (1992) saw language transfer arising whenever “previous knowl-edge” (the L1 in the case of L2 acquisition) constrains the range of hypothesesthe learner will invoke to make sense out of an unfamiliar structure. Whetheror not the notion of hypothesis formation can account for all types of cross-linguistic influence, it does seem applicable here. Moreover, Schachter’s em-phasis on competing hypotheses helps in recognizing that our results indicatemore than one possible source for source-language influence. This latter pointis one that we will return to briefly in our conclusion.

CONCLUSION

The results of our study provide further evidence of the transferability ofbound morphology. More generally, our findings indicate that the coding ofspatial information in the L1 can greatly influence the way such information isrealized in an interlanguage. Whether the information is coded as a preposi-tion, a postposition, or an inflectional morpheme, the L1 can inform the

Page 20: Jarvis & Odlin 2000

554 Scott Jarvis and Terence Odlin

semantic choices a learner makes. It might be tempting to say that the mor-phological coding is irrelevant and that only the semantics matters. That infer-ence would be mistaken, however, given that the performance of the Finnsindicates that when speakers of an agglutinative language such as Finnishmake interlingual identifications with a language such as English, they areprone to overlooking inflectional information equivalent to prepositions in theL2. Although such omissions might seem to be no more than simplification,such instances of zero prepositions are related to the structural contrast be-tween Finnish and English. The fact that different learners choose differentprepositions is also significant, as when some Finns choose to write sit to thegrass whereas others choose sit on the grass. Both choices reflect transfer, neg-ative in the case of the former and positive in the case of the latter. What isespecially significant here is that cross-linguistic influence is not monolithic.In other words, the L1 sometimes offers more than one hypothesis whenlearners look to it for help in acquiring an L2.

(Received 4 November 1999)

NOTES

1. Eubank (personal communication, October 28, 1998) stated that his claims should be under-stood to refer only to the nontransferability of “overt, visible morphology from the NL.” We preferto let his published words speak for themselves, noting that even his claim concerning overt, visiblemorphology is untenable; we return briefly to this issue in the following section of the paper.

2. Finland Swedes prefer not to be referred to as Swedish-speaking Finns, although they are na-tives of Finland whose L1 is Swedish (see, e.g., Ringbom, 1987).

3. Cantonese does have a special class of verb particles that function similarly to English spatialprepositions; however, these particles normally occur either preverbally or postnominally and arerarely used to express directionality (Ball, 1971). In Japanese, spatial relations are expressed throughpostpositions (Kuno, 1978).

4. We should also note that our findings are compatible with the Transfer to Somewhere princi-ple of Andersen (1983). On the one hand, the absence of zero prepositions in the narratives of theSwedes suggests positive transfer from one prepositional language to another. On the other hand,the zero prepositions of the Finns suggest that the difference of morphological type disposes themto overlook the semantic information of Finnish inflections, which are equivalent semantically butnot formally to English prepositions.

REFERENCES

Andersen, R. (1983). Transfer to somewhere. In S. M. Gass & L. Selinker (Eds.), Language transfer inlanguage learning (pp. 177–204). Rowley, MA: Newbury House.

Andersen, R. (1984). The One-to-One Principle of interlanguage construction. Language Learning, 34,77–95.

Ball, J. D. (1971). Cantonese made easy. Taipei, Taiwan: Ch’eng Wen.Becker, A., & Carroll, M. (Eds.). (1997). The acquisition of spatial relations in a second language. Am-

sterdam: Benjamins.Beite, A.-M., Englund, G., Higelin, S., & Hildeman, N.-G. (1963). Swedish grammar. Stockholm: Alm-

qvist & Wiksell.Carroll, M. (1997). The acquisition of English. In A. Becker & M. Carroll (Eds.), The acquisition of

spatial relations in a second language (pp. 35–78). Amsterdam: Benjamins.Comrie, B. (1981). The languages of the Soviet Union. New York: Cambridge University Press.Correa-Beningfield, M. R. (1990). Prototype and second language acquisition. Revue de Phonetique

Appliquee, 95–97, 131–135.

Page 21: Jarvis & Odlin 2000

Morphological Transfer 555

Dawkins, R. M. (1916). Modern Greek in Asia Minor. New York: Cambridge University Press.Dulay, H., & Burt, M. (1974). Errors and strategies in child second language acquisition. TESOL Quar-

terly, 8, 129–136.Dulay, H., Burt, M., & Krashen, S. (1982). Language two. Oxford: Oxford University Press.Ellis, R. (1982). The origins of interlanguage. Applied Linguistics, 3, 207–223.Eubank, L. (1993/1994). On the transfer of parametric values in L2 development. Language Acquisi-

tion, 3, 183–208.Eubank, L., Bischof, J., Huffstutler, A., Leek, P., & West, C. (1997). “Tom eats slowly cooked eggs”:

Thematic-verb raising in L2 knowledge. Language Acquisition, 6, 171–199.Extra, G., & Mittner, M. (Eds.). (1984). Studies in second language acquisition by adult immigrants: Pro-

ceedings of the ESF/AILA Symposium held on 9 August 1984 in Brussels. Tilburg, The Netherlands:Tilburg University.

Harley, B. (1989). Transfer in the written compositions of French immersion students. In H. De-chert & M. Raupach (Eds.), Transfer in language production (pp. 3–19). Norwood, NJ: Ablex.

Homburg, A., & Nikanne, U. (1993). Case and other functional categories in Finnish syntax. Berlin: Mou-ton de Gruyter.

Ijaz, H. (1986). Linguistic and cognitive determinants of lexical acquisition in a second language. Lan-guage Learning, 36, 401–451.

Jarvis, S. (1998). Conceptual transfer in the interlingual lexicon. Bloomington: Indiana University Lin-guistics Club.

Karlsson, F. (1983). Finnish grammar. Juva, Finland: Werner Soderstrom Osakeyhtio.Krashen, S. (1983). Newmark’s “ignorance hypothesis” and current second language acquisition the-

ory. In S. M. Gass & L. Selinker (Eds.), Language transfer in language learning (pp. 135–153). Row-ley, MA: Newbury House.

Kuno, S. (1978). Japanese: A characteristic OV language. In W. Lehmann (Ed.), Syntactic typology: Stud-ies in the phenomenology of language (pp. 57–138). Austin: University of Texas Press.

Laury, R. (1997). Demonstratives in action: The emergence of a definite article in Finnish. Amsterdam:Benjamins.

Martin, M. (1995). The map and the rope: Finnish nominal inflection as a learning target (Jyvaskyla:Studia Philologica Jyvaskylaensia, 38). Jyvaskyla, Finland: University of Jyvaskyla.

Meisel, J. (1980). Linguistic simplification. In S. Felix (Ed.), Second language development: Trends andissues (pp. 13–40). Tubingen, Germany: Gunter Narr.

Meisel, J. (1983). Strategies of second language acquisition: More than one kind of simplification. InR. Andersen (Ed.), Pidginization and creolization as language acquisition (pp. 120–157). Rowley,MA: Newbury House.

Mukattash, L. (1984). Errors made by Arab university students in the use of English prepositions.Glottodidactica, 17, 47–64.

Odlin, T. (1989). Language transfer: Cross-linguistic influence in language learning. New York: Cam-bridge University Press.

Odlin, T. (1990). Word order transfer, metalinguistic awareness, and constraints on foreign languagelearning. In B. VanPatten & J. F. Lee (Eds.), Second language acquisition/foreign language learning(pp. 95–117). Clevedon, UK: Multilingual Matters.

Odlin, T. (1991). Irish English idioms and language transfer. English World-Wide, 12, 175–193.Orr, G. (1987). Aspects of the second language acquisition of Chichewa noun class morphology. Unpub-

lished doctoral dissertation, University of California, Los Angeles.Pavesi, M. (1987). Variability and systematicity in the acquisition of spatial prepositions. In R. Ellis

(Ed.), Second language acquisition in context (pp. 73–82). Englewood Cliffs, NJ: Prentice-Hall.Ringbom, H. (1987). The role of the first language in foreign language learning. Clevedon, UK: Multilin-

gual Matters.Ringbom, H. (1992). On L1 transfer in L2 comprehension and L2 production. Language Learning, 42,

85–112.Rosetti, A. (1945–1949). Langue mixte et melange des langues. [Mixed languages and mixture of lan-

guages.] Acta Linguistica, 5, 73–79.Schachter, J. (1983). A new account of language transfer. In S. M. Gass & L. Selinker (Eds.), Language

transfer in language learning (pp. 98–111). Rowley, MA: Newbury House.Schachter, J. (1992). A new account of language transfer. In S. M. Gass & L. Selinker (Eds.), Language

transfer in language learning (pp. 32–46). Amsterdam: Benjamins.Schumann, J. (1986). Locative and directional expressions in basilang speech. Language Learning, 36,

277–294.Selinker, L. (1992). Rediscovering interlanguage. London: Longman.

Page 22: Jarvis & Odlin 2000

556 Scott Jarvis and Terence Odlin

Sharwood Smith, M. (1979). Strategies, language transfer, and the simulation of the second languagelearner’s mental operations. Language Learning, 29, 345–361.

Sulkala, H. (1996). Finnish as a second language for speakers of related languages. In M. Martin & P.Muikko-Werner (Eds.), Finnish and Estonian: New target languages (pp. 143–158). Jyvaskyla, Fin-land: Center for Applied Language Studies, University of Jyvaskyla.

Thomason, S., & Kaufman, T. (1988). Language contact, creolization, and genetic linguistics. Berkeley,University of California Press.

Weinreich, U. (1953/1968). Languages in contact. The Hague: Mouton de Gruyter.Whitney, W. (1881). On mixture in language. Transactions of the American Philological Association, 12,

5–26.