VERSUS : HOW TO TEACH SECOND LANGUAGE STUDENTS OF By ANN MARIE HEALY A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENT S FOR THE DEGREE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2014
Â© 2014 Ann Marie Healy
To Gainesville, Florida: I escaped to you only to find a new home
4 ACKNOWLEDGMENTS I would like to thank the many peopl e who are integral to my success as a student and as an adult, in general. I thank my parents, my siblings, my academic cohort, my homebase cohort, my amazing students, my former professors, my current professors, and, especially, my committee. Unequivoc ally , I could not have complete d this thesis without their constant support .
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIG URES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 10 ABSTRACT ................................ ................................ ................................ ................... 11 C HAPTER 1 I NT R O DUCTION ................................ ................................ ................................ .... 13 2 L ITERATU RE REVIEW ................................ ................................ .......................... 15 How Learners Learn ................................ ................................ ............................... 15 P edagogical Approaches ................................ ................................ ........................ 23 Prepositions ................................ ................................ ................................ ............ 26 Dans and En ................................ ................................ ................................ ........... 28 Barriers to the Acquisition of Prepositio ns ................................ .............................. 29 3 M E THODOLOGY ................................ ................................ ................................ ... 35 The Context of Acquisition ................................ ................................ ...................... 36 Ã€ vous! and RÃ©seau ................................ ................................ ............................... 36 Justification ................................ ................................ ................................ ............. 42 Participants ................................ ................................ ................................ ............. 44 Materials ................................ ................................ ................................ ................. 45 Procedure ................................ ................................ ................................ ............... 51 Data Analysis ................................ ................................ ................................ .......... 53 4 R ESULTS ................................ ................................ ................................ ............... 54 5 D ISCUSSION ................................ ................................ ................................ ......... 65 Results by Level ................................ ................................ ................................ ..... 65 Results by Treatment ................................ ................................ .............................. 69 Other Factors ................................ ................................ ................................ .......... 75 Implications for Future Research ................................ ................................ ............ 79 6 C ONCLUSION ................................ ................................ ................................ ........ 81 APPENDIX
6 A C OMPLETE PARTICIPANT INFORMAT ION ................................ ......................... 83 B P ARTICIPANT INFORMATION SHEET ................................ ................................ . 85 C T E STING MATERIAL ................................ ................................ ............................. 86 D P EDAGOGICAL INTERVENTIONS ................................ ................................ ........ 89 LIST OF REF ERENCES ................................ ................................ ............................... 96 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 100
7 LIST OF TABLES Table page 2 1 Specialized uses of dans and en with examples in English and French. ............ 28 3 1 Tokens of dans and their linguistic context, with translations (Anover & Antes, 2012, p. 115) ................................ ................................ ................................ ...... 38 3 2 Tokens of dans and en and their linguistic context, with translations (Anover & Antes, 2012, p. 366,7) ................................ ................................ ..................... 39 3 3 Explicit explanations of dans and en in Ã€ vous (Anover & Antes, 2012) ............. 40 3 4 Participant distribution by treatment group ................................ ......................... 45 3 5 Distribution of target tokens in the cloze test ................................ ...................... 48 3 6 Distribution of distractor elements in the cloze test ................................ ............. 49 3 7 Experimental timeline for each participant group ................................ ................ 52 4 1 Pre t est composite data by level and participant group (n=49) ........................... 54 4 2 Post test composite data by level and participant group (n=49) ......................... 54 4 3 Post test results: Composite data analyzed using 2 tailed t tests ....................... 55 4 4 Pre test and post test composite results for target prepositions (n=13) ............. 55 4 5 Pre test composite data for target prepositions, analyzed using 2 tailed t tests ................................ ................................ ................................ .................... 56 4 6 Post test composite data for target prepositions, analyzed using 2 tailed t tests ................................ ................................ ................................ .................... 56 4 7 Pre test and post test composite results for dans (n=7) ................................ ..... 57 4 8 Post test data for dans, analyzed using 2 tailed t tests ................................ ...... 57 4 9 Pre test and post test composite results for target en (n=6) ............................... 58 4 10 Pre test data for en, analyzed using 2 tailed t tests ................................ ............ 58 4 11 Post test data for en, analyzed using 2 tailed t tests ................................ .......... 58 A 1 Participant distribution by gender ................................ ................................ ....... 83 A 2 Participant distribution by age ................................ ................................ ............. 83
8 A 3 Participant distribution by academic major ................................ ......................... 83 A 4 Participant distribution by native language ................................ ......................... 83 A 5 Participant distribution by length of study ................................ ........................... 83 A 6 Particip ant distribution by self reported interest rating ................................ ........ 84 A 7 Participant distribution by self reported knowledge rating ................................ .. 84
9 LIST OF FIGURES Figure page 4 1 Mean aggregate test scores (n=49) ................................ ................................ .... 59 4 2 Mean aggregate scores for combined prepositions (n=13) ................................ . 60 4 3 Mean aggregate scores for dans (n=7) ................................ ............................... 60 4 4 Mean aggregate scores for en (n=6) ................................ ................................ .. 61 C 1 Cloze test, first page ................................ ................................ ........................... 86 C 2 Cloze test, second page ................................ ................................ ..................... 87 C 3 Cloze test, third page ................................ ................................ .......................... 88 D 1 Slide 1, both presentations ................................ ................................ ................. 89 D 2 Slide 2, both presentations ................................ ................................ ................. 89 D 3 Slide 3, both presentations ................................ ................................ ................. 90 D 4 Slide 4, implicit presentation ................................ ................................ ............... 90 D 5 Slide 5, implicit presentation ................................ ................................ ............... 91 D 6 Slide 4, explicit presentation ................................ ................................ ............... 91 D 7 Slide 5, explicit presentation ................................ ................................ ............... 92 D 8 Slide 6, explicit presentation ................................ ................................ ............... 92 D 9 Slide 7, explicit presentation ................................ ................................ ............... 93 D 10 Slide 8, explicit presentation ................................ ................................ ............... 93 D 11 Slide 9, explicit presentation ................................ ................................ ............... 94 D 12 Slide 10, explicit presentation ................................ ................................ ............. 94 D 13 Slide 11, explicit presentation ................................ ................................ ............. 95
10 LIST OF ABBREVIATION S IFE Students in an intermediate level French course who received an explicit pedagogical intervention as part of the present study. IFI Students in an intermediate level French course who receiv ed an implicit pedagogical intervention as part of the present study. NFE Students in a novice level French course who received an explicit pedagogical intervention as part of the present study. NFI Students in a novice level French course who received a n implicit pedagogical intervention as part of the present study.
11 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts By Ann Marie Healy August 2014 Chair: Theresa Antes Major: French and Francophone Studies For a second language (L2) learner, p repositions are one of the mo st difficult fea tures of a language to acquire. This thesis explores the effects of two different kinds of pedagogical intervention on the acquisition of two French prepositions dans and en by novice and intermediate university students of L2 French. These two prepositi ons pose an especially difficult challenge for L2 learners because of their shared meaning in in English but conflicting functions. Novice and intermediate participants were given a multiple choice cloze test designed to solicit their baseline understandin g of dans and en. Following the pre test, half of the students received an explicit pedagogical intervention while the other half received an implicit intervention. These pedagogical approaches were selected for their potentially differing effects on int erlanguage, specifically in regards to the benefits or drawbacks of a rules based approach versus an exemplar based approach. Several weeks after the interventions, the participants completed a second multiple choice cloze test in order to measure their m odified understanding of the target prepositions.
12 The results showed that the presence of an intervention, of either kind, was significantly more beneficial for the intermediate level students than for the novice level students. However, within each partic ipant level, neither type of intervention was proven to be s ignificantly more beneficial tha n the other . The post test results also demonstrated a strong correlation between length of previous exposure to French and success of treatment suggesting the i mp ortance, not only of academic level, but of experience with the language overall.
13 CHAPTER 1 INTRODUCTION Learning a foreign language is universally difficult. There are many linguistic and extralinguistic factors , and each has its own effect on the over all process of learning. Difficult though it may be, it is an important and often necessary part of a university education where many degrees require a certain amount of foreign language instruction. Within this same university system, a certain amount of research is dedicated to understanding how language learning happens and to exploring more effective ways of teaching foreign language. The present thesis is oriented within this framework. Four classes of students at the University of Florida learning F rench as a second language (L2) were tapped to participate in this research regarding the acquisition of prepositions. Prepositions are one of the last language features to be used in a native like manner by L2 speakers. They are part of a very rigid sys tem that is unique to each language and are very rarely addressed directly during classroom time. For these reasons, many L2 users would say that prepositions are among the least developed parts of their L2. Previous research has described the benefits of using spatial relationships to teach prepositions (Boers & Demecheleer, 1998; M. Gulberg, 2010; J. Jones, 2006 ), as well as the benefits of corpus based activities (Fort & Guillaume, 2007; C. Mueller, 2011; Roslim & Mukundan, 2011) and study abroad expe riences (Lafford & Ryan, and whose standard placement in a sentence does not involve the kind of high frequency word pairs that are highlighted in corpus studies. Furt hermore, study abroad experiences aren his study is designed to explore a
14 functional and efficient way to teach complex prepositions in the foreign language classroom . The two target prepositions in this thesis dans and en represent the kind of prepositions that are not directly addressed by current prescriptive research. While they do have a locative function both prepositions translate to in they also have a variety of functions that change depending on the semantic and linguistic context. For example, to select the right kind of locative in , the speaker needs to determine how abstract or concrete th e location being entered is. In this case, simply extending the literal, locative meaning to other figurative context s, as suggested by Boers & Demecheleer (1998), is insufficient to completely describe the complementary but mutually exclusive functions of dans and en. While corpora could be used to solidify certain high frequency examples of dans and en (such as en lign e , meaning online ), these examples may not be sufficient to illustrate the sema n tic differences that selecting dans or en denotes . This is especially true when considering the non locative uses of dans and en which do not generally appear in frequent coll ocates. For these reasons, this study uses two different types of targeted pedagogical intervention implicit and explicit designed to stimulate the development of interlanguage amongst the participants. The participant classes contain both novice high a nd intermediate low students who had limited explicit exposure to these two prepositions prior to the treatment . Using a pre intervention test as a baseline, the post intervention test will help to determine if these interventions are effective and , if th ey are, if one of them is more effective than the other.
15 CHAPTER 2 L ITERATURE R EVIEW This literature review will contextualize the present study. To begin it will orient the reader within the theoretical framework of second language acquisition and giv e information from this field regarding how learners learn language. It will then explore two different pedagogical approaches used to stimulate learning. Following this section, this chapter will address the system of prepositions in French, the specific target prepositions being examined in this study, and will give an explanation for why those prepositions are difficult to acquire. How Learners Learn Second language acquisition (SLA) is a very hotly contested branch of applied linguistics. Due to the s heer number of variables involved, a great deal of SLA research takes place on the theoretical level. Thus, it is important to specify exactly which theories and precepts are being applied when studying a feature of SLA. An important element of these the ories is that they are just that theories. As such, there are often competing ideas within sub fields of SLA. Great effort has been made to be as clear as possible when applying theory to the present study . In general, this thesis is structured around the idea that learners are active in developing their second language. Learners process input for relevant content and that con tent i s stored in one of many systems of language representation in the brain. These systems are intertwined and work in tandem to produce the most efficient and correct language available to the learner at that time. This process can be summarized in the form input and intake, processing and storage, and output.
16 Input is the entirety of language exposure offered to a learner. Clearly, if all input language would be much shorter and much less laborious than we know it to be. There are several theories that describe the mediating factors that pare extensive input down into manageable intake. The present study explores three of them: the Comprehensible Input Hypothesis, Input Processing Theory, and the Noticing Hypothesis. states that input is a necessary and sufficient condition for language learning . Krashen propose d that input only slightly more advanced than the language already acquired is the most effective kind . This hypothesis is symbolized in the formula i+1. When the condition i+1 is met, a learner can reasonably process input for overall meaning . Whether comprehensible input, i+1, is actually both necessary and sufficient for acquisition is the subject of some debate. There are some studies in the field of psycholinguistics that d comprehension, and that the reverse is also true. McLaughlin, Osterhout, & Kim (2004) tested learners of introductory French on their ability to recognize words and non words in French. The researc hers measured the Event Related Potentials of participants following the presentations of an assortment of words and pseudowords. Their data show that, even after a mere 14 hours of instruction at the university level, L2 students are sensitive enough to the rules of French morphology that their brains can recognize a pseudoword (in the form of an N400). However, the participants were not necessarily consciously aware of this and often made judgments that conflicted with their ERP data , wh ich supports Kras .
17 Input Processing Theory (1996), which puts the mediating responsibility on processes performed by the learner that are generally subconscious. According to this theory, learners process input for meaning before they process it for form, which means that they often process open class words before closed class words. By extension, this means that intake tends to include a smaller amount of function words tha n are proportionally available in input and that those function words, which can often have many meanings, might be reduced to having the one meaning that was represented in the input. The third theory that explains the mediation of input into intake is S Noticing Hypothesis (1995), which puts the mediating responsibility directly on the learner. Schmidt proposed that, in order to acquire any feature of language, attention must be paid to the specific feature and that, once noticed, absorption of this feature can begin. Noticing requires a certain degree of intention ality on the part of the learner. As an intentional process, noticing also requires the use of working memory in order to first, remember to notice and secondly, to remember the diffe rent instances of noticing. As such, noticing can be costly in terms of the attentional resources it requires. There are several factors that can affect the cognitive costs of noticing as well as the overall speed with which a feature is assimilated. T hese factors are described by both Schmidt (1995) and Skehan (2001). The first is dependent on the quality of the input. If a feature is especially frequent in the language then there would be increased opportunities for noticing. Noticing would also be increased if that feature were especially prominent in the language. For language features that are not especially
18 frequent or especially salient, instructed input can be quite helpful as it can artificially raise the levels of both . That being said, if noticing is to occur during an instructed task, capacity, a second factor, is described by the Limited Capacity Model of psycholinguistics, which explains that an i ndividual has a limited complement of cognitive resources to be allocated at any given time. If a task is particularly difficult for a learner, much more of their cognitive resources w ill be dedicated to the task tha n they will be to noticing. A third se t of factors that affect the efficacy of noticing is individual differences among learners. Skehan (2001) divides these differences into two items: readiness and notice a certain feature. A linguistic feature that is well beyond the level of language already acquired would require far too many cognitive resources to make noticing feasible. For example, a true novice level learner would not be prepared to notice the functio nal uses of the subjunctive mood in a given second language. They might be prepared to notice verb endings on a regular present tense verb. Differences in individual processing capacity can also affect how noticing happens and how effective it is. Even i f two learners show the same level of demonstrable ability, they may have different capacities to notice due to different individual levels of processing capacity. This means that, over time, certain students will develop their language skills more quickl y than others even if they are presented with equal input at all times. Following input and intake, the second step in the language learning process proposed by this study is processing and storage. Every individual has a model of
19 language in their brain that is unique to them. This individualized representation of language is known as interlanguage. Interlanguage is developed through patterning and formulae. Initially these formulae are vague and general. According to R. Ellis nowledge emerges gradually as learners acquire new sequences, restructure their representation of old sequences, and, over time, extract underlying become more specific and more accurate. Ortega (2009) lists four processes that are common in the development of interlanguage. The first is simplification, which involves a reduction of rules such that there is one form meaning relationship within each language structure. Thi s is common among novice level speakers who use a reduced form of language and often neglect nuances to which more advanced speakers might have access. The second is overgeneralization, wherein one structure or pattern is generalized to other situations . Both of these processes are part of the rule development of interlanguage. Learners over and underuse structures in an attempt to limit and refine the rules governing them. This negotiation can be additive, as when new nuances and layers are added to a process of interlanguage: restructuring. During restructuring rules are changed . Sometimes this necessitates the complete remodeling of a rule based on new information . The remodeling can cause, for a time, what appears to be a reversal in progress as new rules are being built to replace old, faulty ones. For this reason learners often show a pattern of U shaped development (the fourth process). This pattern progress es from relative comfort with an element of language, to a restructuring
20 that causes gaps in performance while new rules are being built, to a renewed mastery of that element once the new rules are in place. This process happens over and over again with m any different features of a language ; some more complex, features require continual restructuring. Skehan (2001), refines the concept of interlanguage into systems of representation that work together to organize and store both rules and patterns. It shoul systems describe what happens as a learner learns; they do not inform the learner of how to learn. A rule based system relies upon the working memory to store rules and uses cognitive resources to determine which of many rules to apply at a given time. These rules are often very specific and several might be applied over the course of a given task. For a novice level learner the rule based system is functional because it is d iscrete and immediately applicable. However, a rule based system can be very taxing on cognitive resources. In the long term, as more and more complex language is used, this means that a rule based system will slow language production due to the massive volume of rules needed to complete a complex task. A second system of representation is an exemplar based system. In an exemplar based system, past input that has been determined to be correct or native like is stored in the long term memory. When new input is received, it is compared to previous exemplars to determine if that input is in keeping with what the learner knows to be correct. An exemplar based system can be less cognitively taxing than a rule based system. However, it is reliant on previo us examples to create language. As
21 such, a lack of exemplar for a certain concept or language convention could lead to difficulty when the learner needs to produce that concept or convention. A major difference between these two systems is the amount of input required in order to put these systems in to practice. A rule based system can be built solely on explicit rules given by an instructor/a textbook/a language program, etc., In this way, very little initial authentic input is required to use a rule based system because it is based on rules and not exemplars. For this reason, beginning learners tend to rely on based system needs exemplars, which means it needs input, and this input is most actionable when it is salient and native like. Learners with more exposure to a language might favor an exemplar received to the production of language. Ideally, these two sys tems will merge at some point to become what Skehan (2001) calls a rule based analytic system. This system applies principals from both previous systems. In this schema, language that is produced using rules can become the exemplars that are used to meas ure future input. In a sense, the learner becomes his or her own source of input. A rule based analytic system is not as cognitively taxing and is not reliant on outside input. However, this system requires that the learner have a well formed complement of rules and exemplars in order to activate it. Once established, a rule based analytic system is stored in the long term memory of a learner and, as a result, can be used with relatively little effort. The typical learner likely has all three systems in place to some degree. This means that linguistic features with adequate rules and exemplars can be produced and manipulated with relative ease
22 while linguistic features lacking sufficient rules or exemplars will be much more difficult to produce. Followi ng production and storage, the final step in the language learning process proposed by this study is output. To explain how output is relevant to acquisition, Swain posited her Pushed Output Hypothesis (Swain, 1985). This hypothesis pinpoints production, not input comprehension, as the catalyst for language acquisition. learner to pay attention to the means of expression needed in order to successfully convey his or her (2010) emphasized the value, not only of pushed output but, also, of collaborative output. Following Vygotsky (1986), Swain describes the Zone of Proximal Development, wherein the language abilities of an individual can be improved in the presence of another when each takes turns being the more knowledgeable party. This collaborative meaning, completes thought, and gives voice to affect. If an individual is not given the opportunity to language, then the power to create meaning, to plan, to attend, to rule based analytic s ystem because that production can serve to exemplify certain language features. It is clear that there are many factors that influence language acquisition even when that process can be simplified into the three steps of input and intake, processing and s torage, and output. In the input phase, only a subset of all possible input becomes intake. This intake must be comprehensible and usually features content words over
23 function words. Novice level learners can tend to neglect the nuanced functions of thes e words in favor of simplified form meaning relationships. In order to intentionally cognitive capacity. For this reason, readiness to notice a feature can increase the efficacy of noticing because it requires less cognitive energy. Once noticed or acquired, an individualized interlingual system stores information about language. This can happen in a rule tate a lot of input, or an exemplar based format, which is less taxing but requires a good deal of input. Once sufficient rules and exemplars governing a language feature are developed, they can be assimilated into a rule based ana lytic system that genera tes it s own input . Such a system is the least cognitively taxing. All of these systems are subject to the processes of interlanguage, and through development and restructuring they are refined. From this point, output can be helpful in developing the ana lytic system because it naturally creates more exemplars that will be available to the learner for the purposes of comparison and rule making. Pedagogical Approaches The field of SLA provides for a distinction between a second language learner, someone wh o is learning a language that they have the opportunity to use outside of the classroom, and a foreign language learner, someone who is learning a language that is not present in their native environment. The participants in this study are all foreign lang uage learners. Furthermore, students being tapped for so called second language research may actually be learning a third or fourth language, and the language in question may have been learned in one of many different contexts. Many languages are acquired outside the classroom, in an immersion setting for example.
24 Naturalistic learning involves maximum exposure with little guided input and can take place both in and outside the classroom. An instructed setting implies a more explicit approach where specif ic information is imparted from the instructor to the students. For the sake of clarity, in this thesis the term SLA should indicate to the reader: acquisition by an adult student in an instructed classroom setting of a language that is at least their sec ond language and is not present in the native environment. When referring to said language, the term L2 will be used. Within an instructed language classroom further distinctions are made regarding how best to deliver information to students. Explicit inst ruction involves a sort of pedagogical full disclosure. The instructor will explain the learning goals, outline and explain all rules involved, and give isolated examples. This closely mirrors the approach taken to second language instruction in the Unit ed States during the majority of the 20 th century. The textbook of choice in an explicit learning environment is a grammar manual. The activity of choice is drilling, wherein students apply the given rules to isolated elements of language. Conjugation d rills are a very common manifestation of explicit instruction. Implicit instruction involves a more global approach. Students are not given learning goals or rules (at least initially) and are exposed to what is known as an input flood. According to Wong (2005): In input flood, the input learners receive is saturated with the form that we the form in any way to draw attention to it nor do we tell learners to pay attention to the f orm. We merely saturate the input with the form. The basic idea here is that by flooding the input with many exemplars of the form, learners will have a n increased chance to notice it (p. 37).
25 In the presence of input, the teacher can then guide student s to deduce rules and usages themselves, using a method known as garden pathing. Implicit instruction, which falls under the pedagogical umbrella that is Communicative Language Teaching, has become increasingly common in modern second language classrooms. T he differences between these two approaches are numerous, as are the difficulties in selecting a pedagogical method that is purely one or purely the other. Even with a bounty of pedagogical approaches available to the foreign language teacher, it is ne arly impossible to customize a standard L2 classroom such that it meets the needs of each student individually. For a student using a rule base d system, explicit instruction may be helpful because it provides the necessary rules and it could save that stu dent from having to parse input int o comprehensible chunks. This could be especially effective for beginning learners because it removes the barrier of cognitive overload. When processing input, beginning learners are often faced with large amounts of la nguage that is beyond their ability to notice and assimilate. The incomprehensible input slows processing to a crawl and makes acquisition very difficult. Hav ing ready made rules may help to circumvent this issue and can stimulate noticing for future inp ut. However, those rules require a good deal of cognitive energy to remember and to execute. For a student using an exemplar based system, implicit instruction can be helpful because it provides exemplars. In this case, enriched input, which highlights the relevant f eatures within the input flood, may be especially helpful because it increases noticing and makes important exemplars more salient. Though this method requires less cognitive energy, it must be active for a longer period of time, giving the learner
26 adequate opportunity to notice the feature in new input and to compare it to their existing exemplars. As with a rule based system, readiness is important for acquisition in an exemplar based system. If the highlighted feature of the input is we ll outside the understanding of language that the learner possesses, it will be difficult for that learner to find similar exemplars in similarly complex input , particularly if those exemplars are not highlighted in the same way. In reality, implicit and explicit approaches are often mixed into an eclectic approach that can accommodate a broader range of learning styles. A combination of implicit and explicit instruction gives the necessary information required for acquisition in both rule based and exem plar based systems. Over time, the learner can store well understood features in a rule based analytic system, which could allow them to dedicate more cognitive resources to processing more complex input. Prepositions Prepositions in French, as in many other languages, serve to establish a relationship between words or groups of words. The nature of this relationship can change depending on the placement of the preposition in the sentence. French prepositions serve two main grammatical masters: nouns and verbs. Where nouns are concerned, prepositions can further describe a noun or be used to separate subjects from objects. Description of a noun can involve designating possession. In English, possession can be indicated by the possessive apostrophe. F or example, in the fragment the apostrophe is used to indicate that the dog belongs to Susan. In French, the preposition de is used to unite the two nouns le chien de Susan . Description of a noun can also involve refining the meaning of that noun. In English, this happens through the use of adjectives. For example, a ring that
27 is gold can be described as a gold, or golden, ring. In French, a preposition is required to unite the main noun with another noun that serves to describe. In this w ay, a noun can be used as an adjective in the presence of a preposition une bague en or means a r subject from its objects by designating their relationship to each oth er. This prepositional function is very similar in both French and English. However, there are many instances where a preposition must be explicitly stated in French that would not necessarily be stated in English. For example, I call my mother implies a directionality of calling from the subject , I , to the object , my mother. In English, this is understood but not stated. In French, a preposition is required: Je tÃ©lÃ©phone Ã ma mere . In regard to verbs, prepositions relate groups of words to verb s by givi ng t hose words a function. These functions answer the questions when, where, how, why, with what goal, with what consequence, and given what conditions. A preposition with a temporal function explains when something happens I left the bar around midnight ; J e suis sorti du bar vers minuit . A preposition with a locative function explains where something happens I ran in the hallway; J dans le couloir . A preposition of manner explains how something happens I responded with anger; J avec colÃ¨re . A preposition of goal explains the reasoning being why something happens I go to my classes to learn; J pour apprendre . A preposition of consequence explains the consequences of an action I yelled such that it frightened him; J pour lui faire peur. A conditional preposition explains the conditions under which a certain event might take place In the event that he comes I will be happy;
28 Dans le cas oÃ¹ il arrive je serai contente . While some prepositions only serve one fu nction, many prepositions serve multiple functions. Dans and En This study examines the acquisition of prepositions by second language (L2) speakers of French. To limit the potential breadth of such a study, two prepositions have been selected for examina tion: dans and en. Both dans and en mean in in English. However, they have very specific usages. The specialized uses of both can be found in Table 2 1. Table 2 1 . Specialized uses of dans and en with examples in English and French. Dans En Duration I ndicates the time before an activity begins Indicates the time before an activity ends English French Je finirai dans 10 minutes. Je finirai en 10 minutes. Time Periods Used before decades Used b efore years, season, and months English French Dans les ann Ã© es Geography Going towards masculine states and provinces Going towards feminine countries, states, and provinces English Fren ch Linguistic Context Requires the presence of a determiner before the noun. Less abstract situations. Requires the absence of a determiner before the noun. More abstract situations. English French Dans le pass En r Ã© alit Dans is used very commonly as an independent preposition. According to Morgenstern and Sekali (1997) it is amongst the first prepositions learned by French children because of its prevalence. For L2 speakers of French dans can also be one of
29 the first acquired prepositions because it has a relatively clean 1 to 1 correlation with the word in in English. All translations of French sentences involving dans into English will include the English preposition in . En is much more co mplex. Though commonly used as an independent preposition, it is also frequently combined with other elements to form prepositional phrases. For example, in front of a present participle en can be used to express manner En en gives the manner in which the cake was eaten. In the example, En en indicate s , at least partially, the time when the telephone conversation took place. En is also combined with other elements to form fixed expressions such as en face de Il agit en snob (H noun: La Russe est en guerre (Russia is at war). It can also be used to indicate the composite materials of a noun: (The silk scarf is beautiful). In ma ny of these cases the use of en in French does not denote the use of in in an equivalent English sentence. This is partially attributable to the fact that French syntax favors the use of prepositional phrases in some cases where English favors an adjectiv e: Il est en colÃ¨re (He is angry). Barriers to the Acquisition of Prepositions The difficulty of acquiring prepositions in an instructed foreign language context has been documented in past studies. Charles Muller (2011) studied the effects of frequency a nd word co occurrence on acquisition of English prepositions by non native
30 speakers. Using a cloze test, Muller demonstrated a significantly higher level of accuracy for high frequency, co occurring word pairs when compared to low frequency, independent pr epositions. Certain prepositions in French occur often as part of collocated fixed expressions, such as Ã and de in verbal expressions. For example, one always calls to someone tÃ©lÃ©phoner Ã one remembers of something se souvenir de . Dans and en are mor e commonly used to locate. That is to say, while dans and en could be found in a predictable word order noun, copula, preposition, noun, for example the nouns they unite will vary according to the needs of the speaker. This lessens the chance that these p repositions will occur regularly in high frequency dans and en are not found in high frequency collocates, the researcher must consider alternative methods to spur the acquisition of independent prepositio ns among L2 learners. The Spanish prepositions por and para share a similar form function distinction to that of dans and en in French. Both por and para translate to English as for. However, as with the target prepositions in this study, they have dist inct and discrete uses. Pinto and Rex (2006) examined the acquisition of these prepositions by L2 Spanish learners in an instructed setting; Lafford and Ryan (1995) studied the same prepositions in a study abroad context. Pinto and Rex (2006) studied nati ve English speaking university students in four different university courses: one first year, two second year, and one third year. In their baseline reading, the students in the first class of year two were more accurate in their use of por and para than any other group, including the third year students. The authors attribute this fact to the repeated emphasis of grammatical elements in years
31 one and two, and an increased focus on literature and culture studies in year three. In accuracy with the different functional uses of por and para the authors state that it is not necessarily fair to present them as a contrastive pair. This approach indicates to learners that if they mean to say for they should use por or para and that if o ne of them is not correct, in a given context, the other must be. The oversimplification leads students to neglect other possibilities, such as null prepositions. The authors suggest a sequential approach to learning por and para, rather than a global one , which could help decrease the cognitive overload of learning both prepositions at once. Lafford and Ryan (1995) examined the use of por and para by university students after a semester abroad in Granada. Their results show that the students did acquire a more native like use of the two prepositions but were the least accurate when using them to establish duration. The authors state that the inaccurate uses of por and para often involved overgeneralization of these two prepositions to places where other prepositions, including null prepositions, would have been more accurate. Both articles warn of the danger that these two to one prepositions seem to pose. If students are focused only on form -por and para both mean for in Spanish in a way analogous to the way that dans and en both mean in in French and not function; it will be difficult for them to acquire the nuanced uses of these prepositions. It is clear from these articles that, absent more specific instruction, students have two tendencies: to o vergeneralize the use of the target prepositions due to their apparent correlation with prepositions in English, and to use the target prepositions interchangeably when there is uncertainty regarding which one to use.
32 Where this study is concerned, the sam e theoretical orientation used to explain how learners learn can also be used to explain why learners have difficulty learning prepositions. Input is necessary for the acquisition of any language. However, learners receive an incredible amount of input o ver the course of their acquisition. For a seasoned L2 speaker this input is almost entirely comprehensible, becomes well integrated into their interlanguage, and can then be called upon for the purposes of production. For a beginning learner, the sheer v olume and complexity of input can be overwhelming. As such, language features that have complex functions, like This can lead to a certain amount of avoidance because, while p repositions are highly frequent, they serve a multitude of functions that would require equally multitudinous rules and exemplars to describe them. Prepositions are not often the subject of instruction and, lacking sufficient rules, it would take an enorm ous amount of input for a learner to be able to notice all the individualized functions of a given preposition. Furthermore, all of this input must be comprehensible for the effective noticing of each function to take place. As detailed information on th e contents of French L2 textbooks will show (see Chapter 3), prepositions are highlighted so rarely compared to content words that students likely do not notice them in the way that Schmidt intended. In fact, the opposite is probable. Students are likely to notice that they are unfamiliar with certain prepositions and, given the paucity of explicit information to orient them, avoid prepositions when possible. This is not to say that closed content words cannot be comprehended, acquired, and produced accu rately in the absence of explicit input. There are many case studies that could be used to refute that claim.
33 However, in a non immersion context such as the one tapped for this study, where directed language input comprises a mere five hours of a given 1 68 hour week (not including homework), there are simply not sufficient opportunities for students to notice, and then acquire, the target language in question . Assuming ample and comprehensible input of prepositions, there are still times when circumvent ing them, in both comprehension and production, can be easier than dedicating the amount of cognitive energy required to learn and use them correctly. According to the Input Processing Theory, overall meaning at the sentence level is often more dependent u pon open class words than closed class words for L2 learners. While closed lexical items, such as articles and prepositions, can help to further refine meaning, they are not usually necessary for a basic understanding of any input. In the event that a pr eposition is essential to comprehension, it s meaning often can be deduced. Or, in the presence of a more knowledgeable speaker, the meaning of a preposition can be negotiated. This is especially true in the case of locative prepositions when simple gestu res can eliminate the need for language altogether. Where production is concerned, prepositions can be circumvented, even during forced output. In the same way that gesturing can function as a stopgap measure when receiving input, it can also be used when idea of a preposition can still be conveyed using approximations. Especially in the case of dans and en understood even if they used that preposition incorrectly. If this were indeed how L2 envisioned. That being said, having to employ alternative methods during production in
34 order to avoid a certai n language feature could certainly stimulate a speaker to try and notice how this feature is used in the future.
35 CHAPTER 3 M ETHODOLOGY In order to assess the acquisition of prepositions specifically dans and en by novice and intermediate level students, the following research hypotheses were developed: H 1 : Novice level students of L2 French will show changed usage of dans and en following a pedagogical intervention. H 2 : Intermediate level students of L2 French will show changed usage of dans and en fo llowing a pedagogical intervention. H 0 : Novice and intermediate level students of L2 French will show no change in their use of dans and en following a pedagogical intervention of any kind. To accurately assess the study data in light of the preceding h ypotheses it is important to mention the assumptions on which this study is based. The first is that students of L2s have difficulty acquiring native like use of prepo sitions. Research cited in Chapter 2 substantiates the claim that prepositions, as clos ed class words, are one of the last grammatical features to be fully assimilated by an L2 speaker. The second assumption is that the participants being tested have skills in line with the skill level of the class. That is to say, it is expected that a st udent in a beginner class is a true beginner and that a student in an intermediate class is truly at the intermediate level. The placement exam policy at the University of Florida allows for this assumption to be made with a good deal of confidence. The t hird assumption is that the results of this study will be generalizable to other similar student populations. The goal of this study is to discern if there is a gap in the current method of French language instruction at UF and, if there is, to determine a reasonable way to fill it. If the information garnered from this study were not applicable to other students it would be a project best not undertaken in the first place.
36 The Context of Acquisition This study was completed in the context of the lower division French program at the University of Florida. In this program the progression of courses is as follows: FRE1130 (novice, first semester), FRE1131 (novice, second semester), FRE2220 (intermediate, first semester), and FRE2221 (intermediate, second semester). There is a review course offered, FRE1134, which covers both FRE1130 and 1131 in one semester. Students will generally take either FRE1130/31 or FRE1134. Normally, students progress linearly through the four main courses. There are some stude nts who enter the program with prior experience and thus start at a higher level. The Department of Languages, Literatures, and Cultures mandates a placement exam, in the form of an SATII subject test, for all new students of French with previous experien ce in the language (high school or community college, for example). Students are placed based on the results of the exam and their prior experience in French. Unless taken over the summer, the 1000 level courses are given five days per week in 50 minute s essions. The 2000 level courses meet four days per week with the same session length. While the teachers of these courses are strongly encouraged to use only French in the classroom, the actual amount used varies from teacher to teacher and thus introduc es a certain amount of variability in the type, frequency, and quality of input received. This variability is acknowledged here but, unfortunately, cannot be controlled for. Ã€ vous! and RÃ©seau The 1000 level of French at UF uses the textbook Ã€ vous! : Th e Global French Experience, hereafter referred to as Ã€ vous!, written by VÃ©ronique Anover and Theresa
37 Antes; now in its second edition. The 2000 level of French uses the textbook RÃ©seau: Communication, IntÃ©gration, Intersections , hereafter referred to as R Ã©seau, written by Jean Marie Schultz and Marie Paule Tranvouez; first edition. It is likely, but not necessary, that students in 2000 level courses have also been exposed to Ã€ vous!, from their 1000 level courses. However, only students at the 2000 level will have experience with RÃ©seau, as it is a more advanced text and, furthermore, being used for the first time in the 2013 2014 academic year. It is important to address the presentation of dans and en in these textbooks as they form the primary foundati on of all introductory French courses at UF. In most cases, the explanations in the textbooks are the most concrete, readily available resources for the students in introductory French. It is not the goal of this thesis to criticize, or even to suggest c hanges to, these textbooks. However, their methodological approach is relevant to the present study. Ã€ vous! begins with several pages of information that help the teacher situate the text in the greater context of l anguage Ã€ vous! And Current Theories of Second Ã€ vous! Does not adopt any one theory of second language acquisition, but instea d borrows concepts from several, incorporating the best of each. We provide ample authentic input, allowing students to deal with both grammar and vocabulary inductively, with a concentration on comprehension, before being asked to use the same structures for production. This same progression is in keeping with cognitive approaches to second language acquisition, allowing students to work with material first as memorized routines or patterns, to be an alyzed and grammaticized later ( Anover & Antes, 2012, p. AIE 13). authentic input implicating targeted vocabulary and grammar, deductive grammar
38 presentations, and skill based activities that target comprehension, production, a nd interpersonal interaction. The preposition dans is first introduced in Chapter 4, Passage 2 1 of Ã€ vous! as part of a greater lesson on locative prepositions. The Passage begins with an input flood of prepositions, situated in a text regarding the orga nization of furniture in the apartment of AurÃ©lie a Swiss personage used to introduce several topics in the textbook. The four paragraph text uses the word dans six times. The tokens and their context can be found in Table 3 1. En is not used in this int roductory text. Table 3 1 . Tokens of dans and their linguistic context, with translations (Anover & Antes, 2012, p. 115) Token Context 1 Ã© In the apartment of Aur Ã© Ã© 2 In the kitchen 3 4 5 6 Following this text there are comprehension questions (p. 116) with two tokens of dans and no tokens of en. At t his point the prepositions are explicitly introduced. In a chart including 13 locative prepositions, dans is given with the English translation in. En is only given as part of a prepositional phrase en face de (across from, facing). There are three exer cises dedicated to this vocabulary; two focused on comprehension and 1 The 'Passage' is a recurrent section in Ã€ vous! that introduces new material implicitly before explicit information is given.
39 one on spoken production. In these activities dans is used 11 times, en is not used at all. Chapter 12 of Ã€ vous! introduces prepositions in relation to geography. The lesson again beg ins with an input flood regarding a hypothetical vacation to Tahiti (implicating both the conditional mood and the prepositions of geography). This text includes two tokens of dans and seven tokens of prepositional en. These tokens can be found, with Engl ish translations, in Table 3 2. Table 3 2 . Tokens of dans and en and their linguistic context, with translations (Anover & Antes, 2012, p. 366,7) Dans En Ã© sie fran Ã§ en hiver en laissant This text and its translation illustrate perfect ly an earlier assertion that dans almost always will translate to in, but that en does so less frequently. The following grammar explanation (p. 374,5) states explicitly that en is to be used in front of feminine countries and continents (and in front of masculine countries towards those places. The three follow up activities all focus on production (two involving text directed production and the other involving open en ded, interpersonal
40 communication). The correct responses to the first two activities involve 11 uses of en, none of dans. The third activity includes no uses of en in the prompts, but does involve two uses of dans. Both prepositions are used frequently throughout the entire textbook but Chapters 4 and 12 are the only chapters to address them explicitly. Table 3 3 shows where and if Ã€ vous explicitly addresses the previously stated functions of dans and en. Table 3 3. Explicit explanations of dans and e n in Ã€ vous (Anover & Antes, 2012) Dans En Duration Indicates the time before an activity begins Indicates the time before an activity ends Explicit Presentation None None Time Periods Used before decades Used before years, season, and months Explicit Presentation None None Geography Going towards masculine states and provinces Going towards feminine countries, states, and provinces Explicit Presentation None p. 374 5 Linguistic Context Requires the presence of a determiner before the noun. Less abs tract situations. Requires the absence of a determiner before the noun. More abstract situations. Explicit Presentation None None In general, Ã€ vous! favors an implicit presentation of these two prepositions. This may be intentional but is more likely a n indirect result of the frequency and function of these two elements in the French language. While it would be difficult to say for certain that all of the above uses are somewhere in the text (but, again, not explicitly addressed), it is highly likely t hat the participants have seen all of these uses at least once. It is, however, highly unlikely that they noticed all of them. It is important to note that at the University of Florida, coverage of Ã€ vous! is split e qually between first semester (C hap ters 1 7) and second semester (C hapters 8 14)
41 1000 level French. A student who has been in both classes has seen both chapters involving explicit information on dans and en. A student who started immediately in second semester French will likely not have rev iewed the information presented in Chapter 4. The introduction to RÃ©seau, the textbook for 2000 level French, does not include are some indications of the methodolog ical approach of the authors. First, they state that RÃ©seau incorporates all levels of the Foreign Language Standards: communication, connections, cultures, comparison, and communities (Schultz & Tranvouez, p. xii). Secondly, they state that RÃ©seau is desi gned to foster critical thinki ng on the part of the students. deficient linguistic skills are in reality the symptom of underdeveloped higher level critical thinking skills within a seco RÃ©seau attemps to address issues of articulation by targeting speaking, reading, and writing skills in ways that systematically foster their development through the encouragement of higher level critical thinking (Schultz & Tranvouez, p. xii xiii). They further state that, while grammar explanations are given in French, they have been reduced to the simplest possible form with supplementary information, PrÃ©cisions, appro priate levels of ability and interest. Finally the authors indicate that the grammar of RÃ©seau is intended to be presented in French using both inductive and input processing approaches. (Schultz & Tranvouez, p. xiii) In terms of explicit presentation, pr epositions are not addressed until Chapter 12 of RÃ©seau , which is beyond the material covered in FRE2220 the intermediate course targeted in the present study (see Methodology, below). Up until that point, the
42 vocabulary given in the text is divided into N oms (nouns), Adjectifs (adjectives), Adverbes (adverbs), and Verbes (verbes). Prepositions are not mentioned. However, the text is replete with examples from francophone literature and media that offer the students access to input involving both dans and en. Again, it cannot be confirmed if the students in FRE2220 are directed to notice the appearance and function of these prepositions in the instances where they do occur. Both Ã€ vous! and RÃ©seau favor implicit uses of prepositions . When presented explici tly (only seen in Ã€ vous! ), only isolated uses of dans and en are discussed. Justification The design and goal of this experiment were selected in an attempt to continue prior research discussed in this literature review. As closed class words, preposi tions are remarkably hard to acquire for an L2 learner. However, among studies addressing closed class words there is relatively little research that addresses prepositions. Prepositions are highly frequent and highly necessary features of a language with which many students struggle for a considerable length of time. If there were an effective way to shorten this struggle the foreign language education community would benefit. Among all the prepositions in French dans and en were selected because they are both quite common and quite difficult for speakers of English to grasp. This is especially true because they often appear to correlate directly with their translated counterparts. Lafford and Ryan (1995) and Pinto and Rex (2006) have shown this to be tru e using por and para in Spanish. The same one to one temptation also exists between French and English with dans , and less so with en. The prepositions also tend to be grouped as a contrastive pair if not dans then en , and vice versa. However, there are other prepositions than can fit into this role as well ( Ã , for example,
43 means in time). This makes their functions even harder to discern, which in turn makes them a good tool to measure the efficacy of a pedagogical intervention. Given the high frequency of dans and en one might expect that learners would assimilate them quickly. However, Muller (2011) showed that high frequency is not as effective on its own as it is when coupled with collocation. Dans and en are very rarely found in high freq uency collocations. These collocations can lead to chunking, which can be used to establish patterns in interlanguage. Without chunking learners must attend to the form and function of these words individually. Furthermore, since the function of these wo rds is closely but not exactly related to English, a great deal of restructuring must take place with each new exposure. To speed this process, two opposing treatments were selected: input based instruction and explicit instruction. R. Ellis (2003) lists these two treatments as part of a set of four options that can be used to promote L2 acquisition (the other two being output based instruction and feedback). They also both play major roles in the history of foreign language instruction. Explicit instru ction circumvents some of the difficulties of interlanguage development because the rules are provided to the student directly. However, as previously stated, complex features require a complex set of rules that can overwhelm a novice level ent interlanguage. On the other hand, while implicit instruction does not directly provide the rules to the learner, input can be structured carefully such that the development of rules can be done using the most salient and relevant input possible. For t heir complementary and yet contrasting effects, these two treatments seemed an interesting and relevant pair to use in this study.
44 Finally, a cloze test was selected in order to test the effect of these treatments. In her article on proficiency assessment standards of SLA, Tremblay (2011) gives several reasons justifying the use of a cloze test as a measure of linguistic competence. First, performance on a cloze test correlates significantly with performance on other standard measures of language proficie ncy, such as the MLACFL and the TOEFL. This their actual knowledge of a given language. Secondly, the kind of scoring method used (exact answer, acceptable answer, a nd multiple choice answer) will produce results similar to those of another scoring method. That is to say, a student taking an exact answer cloze test and a multiple choice cloze test will likely have two very similar scores. Thirdly, cloze tests are pra ctical. They are relatively quickly completed; they can be researcher generated, and their format is flexible enough to test a variety of phenomena. Tremblay cautions, however, that the variability of a cloze test can cause probl ems if not developed corr ectly. Participants The participants in this study were recruited from four classes of French students at the University of Florida: two sections of FRE1131 (Beginning French 2) and two sections of FRE2220 (Intermediate French 1). These courses will be hereafter referred to as NFE and NFI (Novice French Explicit and Novice French Implicit) and as IFE and IFI (Intermediate French Explicit and Intermediate French Implicit). Students were asked to participate on a volunteer basis and all UFIRB protocols wer e followed. The total number of participants recruited was 70. As this experiment took place over the course of several days there were some volunteers who were not always present to participate. Their data will not be subject to analysis . Table 3 4 sho ws the
45 number of participants per class as well as the number of full and partial participants therein. Table 3 4. Participant distribution by treatment group Class Partial Participants Full Participants Total Participants NFE 7 13 20 NFI 4 16 20 IFE 3 14 17 IFI 8 5 13 Total 22 48 70 Four additional full participant students were eliminated from the results analysis as their test scores (either pre test, post test, or both) were beyond the established cutoff of two standard deviations from the mean. Further information gathered from the participant population can be found in Appendix A. Materials The author developed all the materials used in this study. Some raw materials were culled from other sources that will be referenced at the appropriate ti me. The first item was a Participant Information Sheet (Appendix B). This questionnaire was designed to collect supplementary information from participants for potential further analysis. The participants were asked to furnish their age and gender in ord er to examine any possible sociolinguistic implications of the results. Participants were also asked questions regarding their language background. In order to accurately interpret the results this information is very important. Participants with prior, even minimal, experience in French as well as participants who speak another Romance language could have an acquisition pattern that differs from participants with no significant language background. It is important to be aware of these factors. Particip ants were also asked to self report their overall skill level in French and their
46 general interest in learning a second language. While self reporting is objectively unreliable, this data can provide interesting insight into the relationship between motiv ation and acquisition. The second item was the test itself (see Appendix C). This text was borrowed from (Pimsleur & Pimsleur, 1992), a textbook designed as an aid to increase fluency and cultural proficiency. It is staged in three sections . This reading has been introduction, the vocabulary is meant to be simple but used in complex grammatical settings. The grammar included all of the verb tenses and moods covered in Ã€ vous! and RÃ©seau . The text was taken, specifically, from Chapter 27 of the Pimsleur reader, considerations: first, that the text contained many pre existing tokens of dans and en and secondly, that the text would be slightly above the difficulty level normally encountered by novice and intermediate level theory of acquisition. The format of the testing measur e was that of a multiple choice cloze test, wherein words were removed and blanks remained to be filled with one of three options given. This format was selected because it reduced the amount of time taken to answer each question (as opposed to writing a complete word and then worrying over the multiple choice format may be more appropriate for lower level L2 learners than for ipants in this study all constituted lower level students, a multiple choice test seemed the right decision.
47 In terms of the formatting of the measure, several factors were considered. The target elements (converted to testing blanks) were selected prima rily because of their grammatical function in this case, prepositions . The placement of distractor elements was selected to maintain an uncluttered measure. When designating blanks for distractor elements it was important that the measure in its entirety be comprehensible and not visually overwhelming. An effort was made to keep even distance between testing elements to help students maintain momentum but also so that they would continue to read the text throughout the testing period. Having context is necessary in order for a test subject to have the best possible opportunity to answer each question correctly. It was also important that the participants spend equal time on each element in order to avoid possible detection of the intended purpose. For t his reason distractor elements were designed to implicate syntactic and semantic knowledge of the L2. These judgments require more extensive knowledge of the language than morphological judgments primarily gender and number inflection would. A Likert Scal responses became relevant to the study. It is not part of a traditional cloze test. The Pimsleur text originally contained 6 target tokens, which were removed and replaced wi th blanks . Upon rewriting the text, 8 additional tokens were added. The grammatical distribution of these tokens can be found in Table 3 5 Once the target tokens were removed, common prepositions (as well as dans and en ) were given as possible answers. When the blank replaced a target structure, both dans and en were given as possible options (with a third distractor option) to
48 intended. The order of appearance of each op tion was randomized using Excel. Table 3 5. Distribution of target tokens in the cloze test Dans En Duration Indicates the time before an activity begins Indicates the time before an activity ends Tokens 1 1 Time Periods Used before decades Used befo re years, season, and months Tokens 1 3 Geography Going towards masculine states and provinces Going towards feminine countries, states, and provinces Tokens 0 0 Linguistic Context Requires the presence of a determiner before the noun. Less abstract si tuations. Requires the absence of a determiner before the noun. More abstract situations. Tokens 6 3 Total 8 7 Due to a substantial number of incomplete testing measures, the author tallied results from the first two pages only. This resulted in a los s of two target tokens: one dans in front of a noun with a determiner and one en in front of a noun without a determiner. This resulted in seven testable tokens of dans and six testable tokens of en. In addition to the target tokens, the test contained 40 distractor elements. They were selected primarily to maintain an uncomplicated flow of text, i.e., not too many consecutive elements. The distractor elements were all designed to test either semantic understanding of the text or proper use of syntax. W hen possible, simple judgments based on grammatical inflection were avoided. It is important to note that
49 when a preposition was used as a distractor dans and en were never given as options. The distribution of distractors is given in Table 3 6. Table 3 6. Distribution of distractor elements in the cloze test Type Frequency Noun 10 Verb 10 Adjective 10 Preposition 10 The final materials used in this study were two PowerPoints that functioned as pedagogical interventions. One intervention provided an explicit presentation of the grammar rules governing dans and en , while the other followed an implicit approach. Writing this thesis as both a researcher and an educator, it is my personal observation that lower division learners of L2 French do not ha ve native like control of prepositions. The question is: is the current pedagogical method used with these learners , which is predominantly implicit, sufficient to foster their acquisition? It is for this reason that an explicit approach has been suggest ed as an alternative. As previously discussed, the presentations of prepositions in both Ã€ vous! and RÃ©seau err on the implicit side. Ã€ vous! certainly offers more explicit information than RÃ©seau , but this information would not be sufficient to constitut e a fully explicit explanation of dans and en. For this reason, the author considers the implicit PowerPoint to be more of a control condition than a testing condition, although the distinction control and test is not entirely pertinent to the present stu dy. Both presentations began in the same manner: an introduction to the topic, a restatement regarding voluntary participation, and an input flood containing all possible usages of dans and en , meant to ensure equal access to relevant input across both
50 c onditions. However, unlike a classic input flood, the words dans and en appeared in two different colors (distinct from each other and from the main text). As stated by Wong (2005) an input flood is dependent upon the students noticing the critical form. In the context of this study, there was not sufficient time to give ample enough input to virtually assure noticing. As such, the prepositions were highlighted to stimulate noticing in the form of enriched input. The text of the input flood was written by the author on the subject of the musical group Nirvana. It was assumed that students would be more likely to engage in the material if the subject were of some interest to them. While it cannot be confirmed that all participants were even aware of Nir vana, best practice dictates that popular culture is a familiar topic among college students and therefore one that will likely interest them. Following the input flood, the two presentations diverged. The explicit presentation then contained a series o f slides explaining each possible usage of dans and en , with both rules and examples. Following these slides was a practice fill in the blank activity, and finally a slide giving the answers to the activity as well as a justification for each. Participan ts who received this intervention would therefore have been exposed to the words in context as well as discrete rules of their usage, in keeping with the explicit tradition. After the input flood in the implicit presentation, participants were shown a sl ide containing collocations containing dans and en from the input text. In this way, participants were given input with the target structures highlighted and isolated for them in order to increase noticing, but without further explanation. Both presentati ons can be found in Appendix D.
51 Procedure The first step in the study was a visit to classrooms to explain the consent form and to administer the pre test. Students were given only the directions found on the test itself. They were stopped after 15 minut es of test time but were not informed of the time constraints in advance. This research was conducted during regular class time and it was important to respect the time constraints of the professors. As such, 20 minutes five for the consent form and fift een for the test were allocated. The students were not informed of the time limit so as to reduce any pressure they might feel to hurry and knowledge and was thus undesirabl e. Within a week of the pre test, the participants were given a pedagogical intervention. An effort was made to balance the number of participants receiving each treatment. The intervention was given only in French for two reasons: first, to respect the style of the class (both courses are taught exclusively in French) and secondly, to keep the French lexicon of each participant active during the presentation. Switching languages during a task is proven to result in slower reaction times (Misra, Guo, Bo bb, & Kroll, 2012). While the presentation was not a timed task, it was important to remove any distractions that might impede acquisition. No questions were allowed during the presentation in order to control the information each participant received. Fo ur to five weeks following the intervention the students were given a post test, consisting of the original testing protocol. While it may seem initially undesirable to repeat the same measure, this decision was made with several factors in mind. Practic ally, it made sense to administer the same test because it would be extremely difficult to mimic the complexity and the variety of forms evaluated in the pre test. A
52 second measure could have introduced many other variables that would be difficult to cont rol. The length of time between intervention and post test was decided in order to create distance from the pre test and to verify the lasting nature of the intervention. While the participants likely recognized the text, it is highly unlikely that they r emembered the answers they gave to 74 items, especially considering that they were never given the correct answers . In this way, the same measure could be used while the effects of repetition could be reduced. The timeline for each course is given below, in Table 3 7. Table 3 7 . Experimental timeline for each participant group Class Pre Test Intervention Post Test NFE February 18, 2014 Explicit February 25, 2014 March 26, 2014 NFI February 18, 2014 Implicit February 25, 2014 March 26, 2014 IFE Feb ruary 20, 2014 Explicit February 24, 2014 March 17, 2014 IFI February 20, 2014 Implicit February 24, 2014 March 17, 2014 Regarding the process of this study and its relationship to the overall progression of FRE1131: as of the pre test, the participan ts in NFE and NFI had not seen the material contained in Chapter 12 of Ã€ vous! , which was pertinent to their overall performance, as en is introduced in this chapter. The post test was administered the day following the Chapter 12 exam in both NFE and NFI . While one might argue that changes from pre to post test cannot reliably be attributed only to the pedagogical intervention received, as it must be acknowledged that learners focused on en in the classroom during this time as well, learners in NFE and N FI studied this material with
53 the same instructor. One should reasonably expect some change over time due to the classroom instruction received, but any differences between the two novice level participant groups might arguably be attributed to the interve ntion provided by the researcher. As Chapter 4 of Ã€ vous! is covered in FRE1130, it cannot be confirmed that all students in NFE, NFI, IFE, and IFI had seen or reviewed the information regarding prepositions contained therein at any time prior to their pre or post tests. Data Analysis Pre test and post test scores of treatment groups were compared using 2 tailed t tests to determine any significant differences. This analysis was performed in the categories of aggregate test score, combined prepositions sc ore, dans score, and en score. At each stage of analysis, the scores of all four treatment groups were submitted to a one way analysis of variance (ANOVA) . Significant ANOVA results were submitted to post hoc Tukey HSD tests, to determine the source of t he differences. When it was necessary to establish a correlation between two sets of data, Pearson Product Moment Correlations were used.
54 CHAPTER 4 RESULTS To determine if the participants all started at the same skill level a mean pre test score was ca lculated as well as the standard deviation, the range of the pre test scores, and the mean amount of blank responses. As stated in Chapter 3, these figures do not include any information from the third page of the measure. The scores can be found in Table 4 1. Table 4 1. Pre test composite data by level and participant group (n=49) Student Group Mean Score Range Standard Deviation Mean Blank Responses Novice (n= 26) 20 12 30 4.03 4.12 Intermediate (n=18) 22.61 15 30 4.79 4.22 NFE (n=12) 18.5 13 30 4.8 6.92 NFI (n=14) 21.28 18 25 2.81 1.71 IFE (n=13) 21.77 16 30 4.17 4.08 IFI (n=5) 24.8 15 30 6.1 4.6 There were no significant differences detected between the scores of any of the treatment groups. This indicates that, while the novice level student s are technically a course level below the intermediate level students, both groups are performing at the same level according to this measure. Table 4 2. Post test composite data by level and participant group (n=49) Student Group Mean Score Change Range Standard Deviation Mean Blank Responses Novice (n= 26) 20.5 0.5 13 29 4.41 3.81 Intermediate (n=18) 24.33 1.72 19 32 3.58 2 NFE (n=12) 19.4 0.9 13 29 4.83 6.17 NFI (n=14) 21.43 0.15 13 28 3.96 1.79 IFE (n=13) 23.46 1.69 19 32 3.48 1.67 IFI (n=5) 26 .6 1.8 23 30 3.05 3.2
55 These same calculations were performed on the post test results to see if the participants developed, or regressed, similarly over the course of the experiment. A new column has been added to reflect the average change in pre and post test scores. These results can be found in Table 4 2. Although no group made a statistically significant improvement from pre to post test, some statistically significant differences appeared when comparing participant groups . In general, the novice level participants patterned differently from the intermediate level participants, whether considered together or by treatment type. These results can be found in Table 4 3 Table 4 3 . Post test results: Composite data analyzed using 2 tailed t tests Gro up 1 Group 2 p value t value DF Combined Novice Intermediate 0.0039 3.0525 42 By treatment NFE IFE 0.0240 2.4264 23 NFI IFI 0.0080 2.6384 17 An ANOVA indicated a significant difference across the four treatment types, and that this difference was not due to chance (p=0.0080, f=4.522). A post hoc Tukey HSD test indicated a significant difference in the effect of level across the implicit treatment groups NFI and IFI (p<0.05). Table 4 4 . Pre test and post test composite results for target preposition s (n=13) Student Group Mean Pre test Score SD (pre) Range (pre) Mean Post test Score SD (post) Range (post) Change in Mean Score N (n= 26) 5.46 2.00 3 10 5.46 2.04 2 10 0 I (n=18) 6.44 1.79 4 9 8 2.06 3 10 1.56 NFE (n=12) 4.5 1.45 3 7 4.92 1.31 3 7 0.42 NFI (n=14) 6.29 2.09 3 10 5.93 2.46 2 10 0.36 IFE (n=13) 6.31 1.70 4 9 7.62 2.29 3 10 1.32 IFI (n=5) 6.8 2.17 4 9 9 0.71 8 10 2.2
56 Table 4 4 shows the pre and post test results for both target prepositions combined (n=13), including mean scores, sta ndard deviations, ranges, and aggregate change. Considering the results of target prepositions alone, the intermediate group as a whole made a statistically significant improvement in their scores (p=0.0210; t= 2.4199, df=34). The novice level group did not. No other significant improvements were detected . Several significant differences were detected between the scores of participant groups in both the pre and post tests. These results can be found in tables 4.5 and 4.6 Table 4 5 . Pre test composit e data for target prepositions, analyzed using 2 tailed t tests Group 1 Group 2 p value t value DF By level NFI NFE 0.0202 2.4884 24 By treatment NFE IFE 0.0091 2.8493 23 Table 4 6 . Post test composite data for target prepositions , analyzed using 2 t ailed t tests Group 1 Group 2 p value t value DF Combined Novice Intermediate 0.0002 4.0389 42 By treatment NFE IFE 0.0016 3.5705 23 NFI IFI 0.0151 2.7020 17 An ANOVA run on the results of the two target prepositions combined indicated a significant difference across the four treatment groups (p=0.0009, f=6.717). A post hoc Tukey HSD test indicated a significant difference in the effect of level across the implicit treatment groups NFI and IFI (p<0.05) as well as a significant difference in the effe ct of level across the explicit treatment groups NFE and IFE (p<0.05).
57 Table 4 7 shows the pre and post test results for dans (n=7), including mean scores, standard deviations, ranges, and aggregate change. Table 4 7 . Pre test and post test composite r esults for dans (n=7) Student Group Mean Pre test Score StDev (pre) Range (pre) Mean Post test Score StDev (post) Range (post) Change in Mean Score N (n= 26) 2.65 1.41 0 5 2.54 1.56 0 6 0.11 I (n=18) 2.83 1.47 0 5 4.06 1.39 0 6 1.23 NFE (n=12) 2.17 1.5 3 0 5 2 1.13 0 4 0.17 NFI (n=14) 3.07 1.21 1 5 3 1.75 0 6 0.07 IFE (n=13) 2.69 1.49 0 5 3.92 1.61 0 6 1.23 IFI (n=5) 3.2 1.48 1 5 4.4 0.55 4 5 1.2 Considering the results of dans alone, the intermediate group as a whole made a statistically signific ant improvement in their scores (p=0.0146; t= 2.5656, df=34). The novice level group did not. No other significant improvements were detected . While no significant differences were detected between participant groups in the pre test results, several si gnificant differences were detected between the scores of participant groups in the post test results. These results can be found in tables 4 8. Table 4 8 . Post test data for dans , analyzed using 2 tailed t tests Group 1 Group 2 p value t value DF Comb ined Novice Intermediate 0.0019 3.3178 42 By level NFE IFE 0.0022 3.4373 23 An ANOVA indicated significant differences between the four treatment groups where dans alone was concerned (p=0.0053, f=4.918). A post hoc Tukey HSD test indicated a significa nt difference in the effect of level across the explicit treatment groups NFE and IFE (p<0.05).
58 Table 4 9 shows the pre and post test results for en (n=6), including mean scores, standard deviations, ranges, and aggregate change. Considering the results o f en alone, no group made a significant improvement in their scores. Several significant differences were detected between the scores of participant groups in both the pre and post tests. These results can be found in tables 4 10 and 4 11 Table 4 9 . Pr e test and post test composite results for target en (n=6) Student Group Mean Pre test Score StDev (pre) Range (pre) Mean Post test Score StDev (post) Range (post) Change in Mean Score N (n= 26) 2.81 1.49 1 6 2.92 1.35 0 5 0.11 I (n=18) 3.61 1.14 0 5 3.9 4 0.94 2 5 0.33 NFE (n=12) 2.3 1.23 1 5 2.92 1.16 1 5 0.59 NFI (n=14) 3.2 1.63 2 6 2.93 1.54 0 5 0.09 IFE (n=13) 3.62 1.26 0 5 3.69 0.95 2 5 0.07 IFI (n=5) 3.6 0.89 3 5 4.6 0.55 4 5 1 Table 4 10 . Pre test data for en , analyzed using 2 tailed t test s Group 1 Group 2 p value t value DF By level NFE IFE 0.0172 2.5690 23 Table 4 11 . Post test data for en , analyzed using 2 tailed t tests Group 1 Group 2 p value t value DF Combined Novice Intermediate 0.0083 2.7688 23 By level NFI IFI 0.0321 2.333 7 17 An ANOVA indicated significant differences between the four treatment types where en alone was concerned (p=0.032, f=3.245). A post hoc Tukey HSD test indicated a significant difference in the effect of level across the implicit treatment groups NF I and IFI (p<0.05). Figures 4 1 through 4 4 show the changes in score from pre to post test for each treatment group. Figure 4 1 pertains to the aggregate mean test score. Figure 4 -
59 2 pertains to the aggregate mean score of both prepositions combined. F igure 4 3 pertains to the aggregate mean score for dans . Figure 4 4 pertains to the aggregate mean score for en. Figure 4 1 . Mean aggregate test scores (n=49)
60 Figure 4 2 . Mean aggregate scores for combined prepositions (n=13) Figure 4 3 . Mea n aggregate scores for dans (n=7)
61 Figure 4 4 . Mean aggregate scores for en (n=6) From these figures, several general tendencies should be noted. The first is that NFE consistently improved their scores from pre to post test across all measures, exc ept dans alone , where their performance remained unchanged . NFI consistently regressed from pre to post test across all measures, except when scoring the tests as a whole. Both IFE and IFI consistently improved their scores from pre to post test across all measures. However, IFI made much larger improvements than IFE in all measures, except dans alone. While none of these gains or losses w ere determined to be statistically significant, the consistency of the results is certainly suggestive. Regarding the results as a whole, the intermediate participants made significant improvements from pre to post test in the categories of combined prepositions and dans alone. The pre test scores indicated that the novice and intermediate level students started at the same level, as measured by the pre test. The novice and
62 intermediate level students received the same kind of interventions. This improvement suggests that there is an interaction between the interventions and the level of the participants, such tha t the novice level students would not show differing results following an intervention and the intermediate level students would. Interestingly, the post test scores for both dans and en did not reveal a significant difference in the performance of parti cipants receiving different interventions within the same level. That is to say, there were no significant differences found in the post test results of NFE versus NFI and IFE versus IFI. However, there were significant differences in the effects of the s ame treatment across both levels, which are echoed in the results of the post hoc Tukey HSD tests. The post test scores for both prepositions combined showed a significant difference in the scores of the groups receiving an implicit treatment (NFI versus IFI). When dans was considered alone there was a significant difference in the post test scores of the two groups receiving an explicit treatment (NFE versus IFE). When en was considered alone there were two interesting effects. The first is that there was a significant difference in the post test scores of the two groups receiving an implicit treatment (NFI versus IFI). The second was that there was a significant difference in the pre test scores of the two groups receiving the explicit treatment (NFE versus IFE) that was not present in the post test scores. Based on these results , it can be concluded that the first hypothesis novice level students of L2 French will show changed usage of dans and en following a pedagogical intervention has not been ad equately substantiated. The second hypothesis intermediate level students of L2 French will show changed usage of dans and en following a pedagogical intervention has been substantiated in part (the
63 intermediate level students only showed significant cha nge with all prepositions combined and dans alone). Accordingly, the null hypothesis has been tentatively disproven. Analysis of the demographic data provided by the participants suggested that there might also be a relationship between the amount of pr evious exposure to French difference in the length of previous exposure between the novice level and the intermediate level participants. The participants in NFI avera ged 0.75 years of experience with French and the participants in NFE averaged 0.625 years; both groups had a median length of exposure of 0.5 years. The students in IFI averaged 4.4 years of experience with French and the students in IFE averaged 4.12 year s; both groups had a median length of exposure of 5 years. Pearson Product C orrelations were used to determine the relationship between length of exposure and testing performance. In terms of pre test performance, there were no significant correlations between length of exposure and scores in the categories of aggregate pre test, combined prepositions, and dans and en alone. The post test data yielded significant correlations between length of exposure and all tabulated score categories. The aggregate post test score correlated with length of exposure yielded a significant p value of 0.010051 (R=0.3841). The post test scores of combined prepositions correlated with length of exposure yielded a very significant p value of 0.000636 (R=0.495). The post t est scores of dans alone correlated with length of exposure yielded a very significant p value of 0.002879 (R=0.4389). The post test
64 scores of en alone correlated with length of exposure yielded a significant p value of 0.019227 (R=0.3517).
65 CHAPTER 5 D IS CUSSION This chapter will first address the results of Chapter 4 directly. It will then address factors that may have influence d the results study internal factors followed by study external factors with the goal of exploring what the results mean. Final ly, this chapter will offer suggestions for future research. Results by Level The first major result of this study was that novice level students demonstrated no significant changes from pre to post test, regardless of treatment, while the intermediate p articipants displayed some significant changes. There are many potential explanations for this fact. One of the most plausible is the interaction between length of exposure to French and the pedagogical interventions. Due to the strict control of the int roductory French program at the University of Florida, there are almost no students in the second semester of the program who have more than a year of experience in French. Out of the 27 novice level participants, only two claimed to have had more than a year of French experience (2 years and 4 years). Hypothetically, the intermediate level students should have no more than 1.5 years of experience with French. This was absolutely not the case. Of the 18 intermediate participants, only two had the expecte d amount of exposure. The rest had exposure ranging from two to six years. This is an incredibly important difference. It is as though the majority of the intermediate level students had 2 6 years of interlanguage development that was not available to t he novice level students. This variety of exposure could influence the results of the study. Prior to the interventions, there were scant differences in the performances of treatment groups on the pre test. There were no differences between the novice and
66 intermediate groups when they were not separated by treatment type for any pre test measure. Additionally, the Pearson product correlations showed no significant relationship between length of exposure and pre test scores. That is to say that, regardl ess of the differences in exposure, the demonstrable ability of each group was relatively similar at the time of the pre test. Following the interventions, several significant differences appeared between the novice and intermediate groups and between dif ferent treatment groups. Furthermore, all of the post test measures (aggregate test score, combined prepositions, dans alone, and en alone) correlated significantly with length of exposure. Without any intervention, the novice and intermediate level stu dents would likely have continued to perform at the same level. With the interventions, however, the intermediate level students were able to pull ahead significantly, relative to the novice level students. This result likely has two contributing factor s: the internal language systems in which dans and en were processed and the effects of restructuring. As discussed in Chapter 2, beginning learners tend to demonstrate usage of a rule based language system. This system is reliant on working memory and c an be costly in terms of cognitive resources. In this experimental design, a rule based system would be at a disadvantage because of the amount of downtime between the intervention and the post test. During this four to five week period, students using a rule based system would have to keep those rules fresh in their minds for the post test, absent any additional input related to dans and en . For the novice level students receiving an explicit treatment, those rules were provided clearly. For the implic it treatment, the process of developing rules from in the input was the responsibility of the students. In
67 memories in order for them to exploit those rules on the post test. A student with more exposure to French might be using an exemplar based system (increased exposure means increased opportunities to collect native like exemplars), or even a rule based analytic system, which combines both rules and exemplars. In the case of both, exemplars and rules are stored in long term memory and the presence of new input causes comparisons between that input and the exemplars already in mind. In this case, the student would not be reliant on the input of the intervention to be a source of rules or exemplars. Instead, that input would confirm or refine those exemplars leading to an enhanced understanding of the material, which would be evident in the post test. Given this information, it is possible that the initial pre tes t results showed no differences between novice and intermediate level students because there were no differences between them. That is to say, the novice level students provided a baseline reading of what classroom exposure had taught them in regards to dans and en. The intermediate level students, despite having at least an additional semester of French under their belt, had not received any additional training in dans and en during that semester. What the intermediate level students did receive, howeve r, was more general exposure (in some cases quite a lot of exposure compared to the novice level students). For the novice level students, the intervention likely signaled a gap in knowledge and, for the novice explicit students, provided specific rules t o meet that gap. For the intermediate level students, who had an additional 0.5 5.5 years of exposure, the interventions could have started the comparison process seen in both exemplar -
68 based and rule based analytic systems. In essence, it is possible tha t the interventions instigated the integration of dans and en into a rule based system, for those already using one, whereas the interventions would have provided more salient, concise examples to confirm the exemplars already present in an exemplar based system. This might account for the apparent similarity in scores prior to interventions and the significant differences that appeared following them, not just as an effect of treatment but also as an effect of exposure. Among the novice level participants there were no significant changes detected in the relationship between the pre and post test scores in general, between the pre and post test scoring of both prepositions combined, between the pre and post test scoring of dans , and between the pre and post test scoring of en . In some cases, the novice level students even appeared to regress (although this regression was not statistically substantiated). This can be attributed to language systems but also to the process of developing interlanguage as a whole. As discussed in Chapter 2, one of the main processes of interlanguage is restructuring. During this process, rules can be modified through both addition and subtraction. It is likely that the effect of treatment on the novice level participants might not be immediately detectable because those participants are in the process of restructuring. If the rules given in the treatments were at odds with the existing rules, one could even expect to see a drop in progress following the treatment. This r egression leads to the U shaped development that is characteristic of interlanguage. Without knowing exactly where on the U shape the post test fell it is difficult to conclude with any certainty the effects of the treatments as a whole, and of the treatm ents individually, on the novice level participants. Furthermore,
69 a lack of significant improvement should not be conflated with a lack of learning. For the present study, it can only be concluded that, at the moment of the post test, any potential effects of intervention were not yet detectable. This could be attributed to restructuring or to the fact that a rule based system would then require sufficient exemplars in order for dans and en to be integrated into long term memory, rather than working memory . A longitudinal approach could be used to determine the ultimate influence of the treatments on the use of dans and en by novice level students. Results by Treatment The next major finding of this study is the apparent insignificance of treatment withi n level, that is to say NFE versus NFI and IFE versus IFI. The results based on treatment groups (NFE, NFI, IFE, IFI) show mixed advantages of each treatment for each level. However, the treatment groups, especially at the intermediate level, were quite sm all and may have lacked sufficient data to generate statistical significance. In general, it appears that the explicit intervention was slightly more effective for novice level students and the implicit intervention was slightly more effective for intermed iate level students (see Figure 4 2). Looking at both prepositions combined, there was a significant difference in the scores of NFE versus NFI, with NFI outscoring NFE by 1.79 points, at the pre test. This difference was neutralized in the post test res ults; the students in NFI appear to have regressed while the students in NFE appear to have advanced. In this case, the significance of treatment may not have been detectable because a significant difference was already present at the pre test. This is t he only example from the data that clearly indicates a difference in the efficacy of treatment within level.
70 Although differences were not confirmed between different treatments in the same level, a significant difference did appear when comparing the pre and post test scores of each treatment across both levels. When all prepositions were scored in combination there was a significant difference in the results of NFI versus IFI and NFE versus IFE (this specific difference was also present in the pre test) . Given that the intermediate level students made a significant improvement in this category, where the novice level students did not, it can be concluded that the treatments were more effective for the intermediate level students than they were for the n ovice level students. The post hoc Tukey HSD test confirmed a significant difference in the effects of both treatments (implicit and explicit) across both levels (novice and intermediate). If it were assumed that a novice level student favored a rule ba sed system, it follows that the intervention providing rules would be more helpful. If the intermediate level students had sufficient exposure to support an exemplar based or rule based analytic system, it follows that providing more exemplars would be mo re helpful. It should be noted here that, while the implicit PowerPoint was very true to its intended form, the explicit PowerPoint was not. The intervention did provide all the necessary rules but it also gave the students access to an input flood. In t his way, the treatment an eclectic approach. This could have been helpful to the novice level students who may have been consciously collecting rules and potentially stor ing exemplars for future comparison . When dans was scored alone, there was a significant difference in the performance of NFE and IFE. The post hoc Tukey HSD test confirmed this difference in
71 effect of explicit treatment. Again, given that the intermed iate level students made a significant improvement in this category, where the novice level students did not, it appears that the explicit treatment was more effective for the intermediate level students than it was for the novice level students, while the implicit treatment was not. However, looking at the results of Table 4 7, we see that the both groups of novice level students regressed similarly (NFE: 0.17, NFI: 0.07) while both groups of intermediate level students progressed similarly (IFE: +1.23, IFI: +1.2). It happens to be the case that this regression and progression was slightly more pronounced for the explicit groups. These slight differences were sufficient enough to generate significance . From a descriptive standpoint, it can be said that the treatments were more effective, in general, for the intermediate level students than for the novice level students, and that this effect is so pronounced as to be significant when examining the explicit treatment groups specifically. Overall, neither the novice nor the intermediate level students showed a significant change in use of en following the pedagogical interventions. This is the only scored category where the intermediate level students did not show a significant improvement. Chapter 2 con tains information regarding the potential preference for dans by students with English as an L1, stating that it has a much more direct relationship with the English word in than en does. Furthermore, dans is presented before en in the textbook Ã€ vous! , t o which the majority of all participants had been exposed. Given these two facts, it would be reasonable to assert that most of the participants, regardless of level, were more comfortable with dans than with en . If this is the case, then the same lack o f preparedness and familiarity that could have hindered
72 the novice level students generally could also have hindered the intermediate level students where en was concerned. In regards to the novice level students , Table 4 9 shows that the students in NFE improved their score from pre test to post test, which is not consistent with the results found for dans (where both novice level groups appear to have regressed). As discussed in Chapter 3, the novice level students (both explicit and implicit) had taken a chapter exam the day before the post test that contained an exercise implicating en . In order to prepare for this exam, the novice level students would have had some additional practice with en that the intermediate level students did not, and this pra ctice would have taken place after the pedagogical interventions highlighted the importance of en in relation to this experiment. Some interesting results emerged due to this unique difference in preparedness. To begin, the pre test scoring of en showed a significant difference in the scores of NFE and IFE, where NFE scored significantly lower than IFE. At the post test this effect had been mitigated. While NFE did still score lower than IFE, the difference in scores was no longer significant. This re sult can be approached in two different ways: the explicit intervention may have been more effective for the novice level participants, helping them close the gap in scores, or the explicit treatment may have been detrimental (in the short term) to the int ermediate participants, causing them to lose ground. The explanation likely lies somewhere between these two approaches. If the novice level students rely on a rule based language system, they are likely to be more susceptible to explicit interventions ( as compared to implicit interventions) because the
73 rule making has been done for them. Intermediate level students would not necessarily favor an explicit intervention because their interlanguage might developed such that they would benefit more from exem plars than from rules hence, why the students in IFI made improvements in their scores consistent with those found when scoring dans . This possible increased efficacy of explicit interventions among novice level students is not evident in other areas of t he study because the explicit intervention marked only the initiation of interlanguage change. This first change to interlanguage would be the most drastic since the interventions contained only target like rules that would not need to be restructured lat er. However, in the case of en, the novice level students were exposed to both an intervention and then explicit information in class. It is possible that this repeated intervening sped the novice level students through interlanguage development and provi ded them with exemplars to use for comparison in the post test. The results as a whole indicate several general tendencies. First, no change was detected among novice level students as a result of either intervention. This change could appear later if the interventions caused restructuring of interlanguage, which would delay results. For this reason, it cannot be asserted that either intervention was more beneficial for the novice level students than the other. However, the results relating to en sugg est that novice level students might show significant benefits from an explicit intervention if that intervention were repeated regularly. Secondly, the intermediate participants as a whole did show significant improvement with dans and with both prepos itions combined. However, when the intermediate participants were divided intro treatment groups (IFE and IFI), the significance of the improvement disappeared. This indicates two things: first, that
74 neither intervention can be proven to be the cause of the overall improvement and, secondly, that the data from both treatment groups were necessary in order to prove significant change. In short, while the interventions did cause significant change, this change cannot be attributed to only one or the other. Some significant results emerged when comparing the post test scores for the same treatment across both levels. For both prepositions combined, this difference appeared when comparing the implicit treatment groups (NFI and IFI) and the explicit treatment groups (NFE and IFE); for dans this difference appeared when comparing the explicit treatment groups (NFE and IFE); for en this difference appeared when comparing the implicit treatment groups (but may also have appeared when comparing the explicit treatm cannot be determined conclusively if these differences resulted from regression of one group, improvement of another, or both. As the intermediate participants did make significant i mprovement in their scores, it is tempting to say that the treatments helped them. But, since no single treatment was identified to be more helpful than the other, it is difficult to attribute a difference in score to the beneficial nature of a specific i ntervention. Likewise, as the novice level participants showed no significant change in results for any measure, it is difficult to attribute a difference in score to the detrimental nature of a specific intervention. Descriptively, it appears that the interventions were generally unhelpful for the novice level students, but in the event that one treatment was more effective than the other it would be the explicit treatment. The novice level students may be working on a rule based system. As such, a tre atment that aids in the development of those rules
75 might help speed acquisition. It also appears that both interventions were generally helpful for the intermediate level students, and in the event that one treatment was more effective than the other it w ould be the implicit treatment. The intermediate level students may be working on an exemplar based system, or even a rule based analytic system. If this is the case, the exemplars provided in enriched input could help hone the system of exemplars being used to compare and generate new language. For the purposes of this study, the author concludes that the effect of treatment across both levels is worth examining but cannot be determined conclusively using the results of the present experiment. Finally , there is clearly a relationship between length of exposure and efficacy of interventions. There may be some evidence to suggest that the effect of treatment be that the novice level students had relatively little exposure while the intermediate level students had had a good deal of exposure. If it is, in fact, the case that these results show an effect of exposure and not of level, that would explain why the pre test res ults showed no difference between novice level and intermediate level students. however, increase the readiness of a student to receive an intervention and to turn th at intervention into actionable information. In this way, there is no correlation between pre test scores and exposure and there are significant correlations between post test scores and exposure. Other Factors As an extension of these results, there are several factors that may have influenced the outcome of the study. The length and difficulty of the testing measure are
76 factors internal to the study that could have influenced the results. The text of the testing measure was selected to meet the requirem Model (see Chapter 2). However, a look at the pre test scores reveals that this may not have been borne out. The students in the novice level group performed slightly above chance and the highest scoring intermediat e group scored just above 50% accuracy. Considering that the target prepositions consisted of only 13 out of 49 testing items, it is clear that the distractor elements were quite distracting. If students have a limited amount of cognitive energy to devot e to the testing measure, and the distractor elements were both difficult and more numerous that the target questions, it is not surprising that there was relatively little improvement from pre test to post test, even after the students were induced to not ice the target prepositions. That being said, the participants did not receive a treatment of any kind regarding the distractor elements. As such, it would not be entirely reasonable to expect a large improvement in the post test given no relevant interv ention . It is, however, important to note that the difficulty of the measure could potentially have favored the intermediate level students over the novice level students. Given the apparent difficulty, and thus cognitive weight, of the measure, it is po ssible that the results regarding dans and en were not representative of the actual knowledge possessed by each participant. However, these prepositions do not usually appear in insolation during normal language use. Therefore, these results, which could have been influenced by the difficulty of the measure, are likely the best representation dans and en.
77 The scores taken from participant 4594 serve as excellent anecdotal evidence of the relationship be tween the cognitive load of the distractor elements and accuracy with the target items. This participant was an intermediate level student who received the implicit treatment. Their overall score on the pre test was a 13 out of 49, which was considerably worse than average. On the post test, 4594 only answered the questions related to dans and en . For the testing items, their score from pre to post test increased by 7 (out of 13). This is a dramatic increase that no other participant in the study was a ble to muster. However, they did not answer a single distractor question. They were able to dedicate their cognitive resources to one topic, which benefited greatly, but were required to ignore the other elements as a result. As such, they were omitted from the study. Regarding study external factors, there are many qualities that each participant brings to a study that can have a large influence on its outcome. Among these qualities are aptitude, motivation, personality, and memory. The aptitude of a student can determine how accurately and quickly that student absorbs and retains new information. Clearly, high aptitude students are more likely to acquire and reproduce the information shared in either type of intervention than would a low aptitude stu dent. This study contained no measure of aptitude outside of the testing measure itself. However, it should be considered that, outside of a French and Francophone Studies major or minor , students are not required to take any language course past FRE1131 (the novice level). Only students desirous of additional foreign language education continue, and these students, in general, have p er formed well in prior courses. Therefore, it is likely that there is a higher proportion of high aptitude students in t he intermediate courses
78 than in the novice level courses. This statement cannot be proved at the current time but is worth considering when examining the results as a whole. Motivation and personality can also have a huge influence on the outcome of an SL A study. Following the pre tests in each class, many participants asked the researcher directly about the goal of the study. While these questions were left unanswered, this was a clue to the fact that some students were more interested in the study as a whole than were others. After the intervention, it would have been clear what the focus of the study was. A highly motivated student may have done supplemental research if they found the intervention to be insufficient. This extra research would have u ndoubtedly affected their overall score on the post test. However, this effect could have been positive, given clear and helpful information, or negative, given dense and confusing information. On the other hand, several participants were noticeably bore d when completing the testing measure. While the cloze test did originally include a Likert Scale to determine certainty of response, many students filled in the scale as a last step, after having responded to all the questions, and many of them circled e xclusively indications from certain participants that they did not give th eir complete attention to the measure, such as occasional text messaging, lengthy pauses, and incomplete measures with time remaining. Some participants were eliminated from the study due to overall scores well outside of the mean (see Chapter 3) but it i s impossible to know if students who fell within the acceptable range were actually performing to the best of their abilities.
79 Memory is another important factor to consider. By design, the post test was given several weeks following the pedagogical inte rventions (ranging from three to four weeks). Students receiving an explicit intervention could have interpreted this intervention as a list of rules to be memorized (rather than inferring the correct usage, as the implicit students may have done). Given even the best of intentions, without further noticing or practice it is possible that the effect of the intervention was too short lived to be noticeable. In the event that the intervention had been done earlier, it is possible that the students receiving an explicit intervention would have performed better, but that better performance would not necessarily have been indicative of actual acquisition. Furthermore, it is difficult to quantify the effect of the relationship between memory (which could have be nefitted from an earlier post test) and interlanguage (which could have benefitted from a later post test). Implications for Future Research For future research, the author has some suggestions to improve this paradigm . The first is a stricter set of limi tations governing the participant pool. The participants tapped for this study were chosen based on necessity and convenience. While their results provided some interesting information, the inconsistencies among them could easily have obscured some of th e results. It would be quite interesting to find students who tested into the same level but have different amounts of previous experience with French. In this way, it would be much easier to determine the effect of treatment in relation to exposure and to avoid potentially conflating exposure with level, as may have happened in this study. Secondly, a shorter and less difficult measure is recommended. It would still be important to base the distractor elements in syntax and semantics rather than
80 morphol ogy, because morphological judgments only implicate a limited set of skills, but the author recommends using a less complex vocabulary overall. It might even be useful to select a sample text from the actual textbooks being used in class. Selection of a text from a slightly more difficult chapter would still push the students but ensure A revamp of the interventions is also recommended. Following Pinto and Rex (2006), it would be advantag eous to include a text in English that is translated into French as a part of the intervention. This achieves the purpose of showing that not all instances of in in English will become a dans or an en in French. It also shows the students that dans and e n are, in fact, not equivalent in function and do have differing uses. A text that is already in French, consistent with the present interventions, would not necessarily convey these ideas. It would also be helpful to eliminate the input flood from the e xplicit presentation. This would certainly help to distinguish between the efficacies of the implicit treatment compared to the explicit treatment. Finally, it could be beneficial to try repeated training sessions rather than one intervention. So few ele ments of a language are fully understood after one exposure, even when the participants are forced to notice it. It is not necessarily reasonable to expect assimilation of a feature after only one exposure, especially when considering the U shaped develop ment of interlanguage. Pinto and Rex (2006) suggested a sequential approach to repeated training. In this way, the interlanguage for each preposition could be built without respect to the other. This sequential training could reduce potential interferenc e in the form of conflation.
81 CHAPTER 6 CONCLUSION As stated in Chapter 2, closed class words are the last to be fully assimilated by a foreign language learner. This is for many reasons, including the particularities of their functions and their potent ial to be avoided when negotiating for meaning. The goal of this thesis was to determine if there was a way to teach certain closed class words to speed up acquisition. Two prepositions were selected, dans and en , which are generally equivalent to in in English. As with por and para in Spanish, foreign language students tend to see these prepositions as being functionally equivalent and as having a one to one correlation with their English counterpart. For these reasons, these two prepositions are espec ially difficult to acquire . These prepositions were the subject of two types of pedagogical intervention. The explicit intervention offered input and rules of usage while the implicit intervention baseline understanding was measured with a pre test and their modified understanding was measured with a post test following the interventions. The results of this study do not prove a statistically significant difference in the effect of the interventio ns. However, they do suggest a strong interaction between level and intervention, which could actually be an interaction between length of exposure to French and intervention. Intermediate level students did benefit from interventions of both kinds in a w ay that the novice level students did not (at least when measured at the time of the post test). It seems that interventions regarding prepositions may not be necessary at the novice level, given that the immediate benefits are significantly larger
82 once t hose novice level students have further developed their interlanguage at the intermediate level. This study has provided an interesting basis for future research that could refine many of the conclusions made . Further information, specifically regarding the benefits or drawbacks of each type of intervention and their relationship to length of exposure would be helpful in suggesting improvement to pedagogical practice.
83 APPENDIX A COMPLETE P ARTICIPANT I NFORMATION Table A 1 . Participant distribution by gender Gender N % Female 37 84 Male 7 16 Other 0 0 Total 44 Table A 2 . Participant distribution by age Age N % 17 19 26 59 20 22 17 39 > 23 1 2 Total 44 Table A 3 . Participant distribution by academic major Major N % Liberal Arts 23 52 S ciences 15 34 Business 5 11 Exchange Student 1 2 Total 44 Table A 4 . Participant distribution by native language Native Language N % English 39 89 Romance Language 2 4 Other 3 7 Total 44 Table A 5 . Participant distribution by length of study Length of Study N & <1 year 25 58 1 <2 years 0 0 2 <3 years 2 5 >3 years 16 37 Total 43
84 Table A 6 . Participant distribution by self reported interest rating Interest Rating N % Very Low 0 0 Low 1 2 Average 10 23 High 19 43 Very High 14 32 Total 44 Table A 7 . Participant distribution by self reported knowledge rating Knowledge Rating N % Very Low 0 0 Low 9 20 Average 32 73 High 3 7 Very High 0 0 Total 44
85 APPENDIX B PARTICIPANT INFORMAT ION SHEET Participant ID Number:_____ _____ Participant Information Sheet 1. Gender:__________________________ 2. Age:___________________________ 3. Major:___________________________ 4. Minor:_________________________ 4. What language(s) do you use most often at home:________ _____________________ 5. Are you currently taking courses in another language (not English or French): ________________________________________________________________________ 6. Have you taken courses in another language in the past: _________________ _______________________________________________________ 7. What French course(s) are you currently taking: ________________________________________________________________________ 8. For how long have you been studying French: _________________________ _______________________________________________ 9. In what context (high school, college, etc.,): ________________________________________________________________________ 10. How would you characterize your interest in learning French? (circle one) Ve ry Low Low Average High Very High 11. How would you characterize your knowledge of the French language? (circle one) Very Low Low Average High Very High
86 APPENDIX C TESTING MATERIAL Figure C 1 . Cloze test, first page
87 Figure C 2 . Cloze t est, second page
88 Figure C 3 . Cloze test, third page
89 APPENDIX D PEDAGOGICAL INTERVEN TIONS Figure D 1 . Slide 1, both presentations Figure D 2 . Slide 2, both presentations
90 Figure D 3 . Slide 3, both presentations Figure D 4 . Slide 4, implicit presentation
91 Figure D 5 . Slide 5, implicit presentation Figure D 6 . Slide 4, explicit presentation
92 Figure D 7 . Slide 5, explicit presentation Figure D 8 . Slide 6, explicit presentation
93 Figure D 9 . Slide 7, explicit presentation Fig ure D 10 . Slide 8, explicit presentation
94 Figure D 11 . Slide 9, explicit presentation Figure D 12 . Slide 10, explicit presentation
95 Figure D 13 . Slide 11, explicit presentation
96 LIST OF REFERENCES Anover, V., & Antes, T. (2012). Ã€ Vous!: The g lobal french experience (2nd ed.). Boston, MA: Cengage Learning. Bley Vroman, R. (1990). The logical problem of foreign language learning. Linguistic Analysis, 20 , 3 49. Boers, F., & Demecheleer, M. (1998). A cognitive semantic approach to teaching prepo sitions. ELT Journal, 52 (3), 197 204. Brossier, E. (Ed.). (2006). Bescherelle: La grammaire pour tous . Paris: Hatier. Brown, J. D. (1980). Relative merits of four methods for scoring cloze tests. Modern Language Journal, 64 (3), 311 317. Brown, J. D. (19 86). Cloze procedure: A tool for teaching reading. TESOL Newsl., 20 (5), 1. Ellis, R. (1999). Input based approaches to teaching grammar: A review of classroom oriented research. Annual Review of Applied Linguistics, 19 , 64 80. doi:10.1017/S026719059919004 4 Ellis, R. (2005). Measuring implicit and explicit knowledge of a second language : A psychometric study. Studies in Second Language Acquisition, 27 (02), 141 172. doi:10.1017/S0272263105050096 Fort, K., & Guillaume, B. (2007). PrepLex: A lexicon of frenc h prepositions for parsing. SigSem '07 Proceedings of the Fourth ACL SIGSEM Workshop on Prepositions, 17 24. Gullberg, M. (2010). Methodological reflections on gesture analysis in second language acquisition and bilingualism research. Second Language Rese arch, 26 (1), 75 102. doi:10.1177/0267658309337639 Hahne, A. (2001). What's different in second language processing? evidence from event related brain potentials. Journal of Psycholinguistic Research, 30 , 251 66. Henry, N., Culman, H., & VanPatten, B. (20 09). More on the effects of explicit information in instructed sla. Studies in Second Language Acquisition, 31 (04), 559. doi:10.1017/S0272263109990027 Hickmann, M., & Hendriks, H. (2006). Static and dynamic location in french and in english. First Languag e, 26 (1), 103 135. doi:10.1177/0142723706060743 Jones, J. (2006, June 2). Picturing prepositions. The Times Educational Supplement
97 Katz, S. L., & Blythe, C. S. (2007). Teaching french grammar in context: Theory and practice . New Haven, MA: Yale Universit y Press. Knudsen, E. I. (2004). Sensitive periods in the development of the brain and behavior. Journal of Cognitive Neurscience, (16), 1412 25. Krashen, S. (1985). The input hypothesis . London: Longman. Lafford, B. A., & Ryan, J. M. (1995). The acquisi tion of lexical meaning in a study abroad context: The spanish prepositions por and para. Hispania, 78 (3), 528 547. Lapkin, S., Swain, M., & Psyllakis, P. (2010). The role of languaging in creating zones of proximal development (ZPDs): A long term care re sident interacts with a researcher. Canadian Journal on Aging/La Revue Canadienne Du Vieillissement, 29 (04), 477 490. doi:10.1017/S0714980810000644 McLaughlin, J., Osterhaut, L., & Kim, A. (2004). Neural correlates of second language word learning: Minima l instruction produces rapid change. Nature Neuroscience, 7 , 703 704. Misra, M., Guo, T., Bobb, S. C., & Kroll, J. F. (2012). When bilinguals choose a single word to speak: Electrophysiological evidence for inhibition of the native language. Journal of Me mory and Language, 67 , 224 237. Morgenstern, A., & Sekali, M. (1997). L'acquisition des premiÃ¨res prÃ©positions chez un enfant francophone. Faits De Langues: Revue De Linguistique, 9 , 201 210. ions: Collocational knowledge or knowledge based on meaning? System, 39 (4), 480 490. doi:10.1016/j.system.2011.10.012 Mukundan, J., & Roslim, N. (2009). Textbook representation of prepositions. English Language Teaching, 2 , 13+. Odlin, T. (2003). Cross l inguistic influence. In C. J. Doughty, & M. H. Long (Eds.), Handbook of second language acquisiton (pp. 436 86). Malden, MA: Blackwell. Ortega, L. (2009). In Comrie B., Corbett G. (Eds.), Understanding second language acquisiton (1st ed.). London: Hodder Education. Osterhout, e. a. (2002). Words in the brain: Lexical determinants of word induced brain activity. Journal of Neurolinguistics, 15 , 171 87. Perani, D., & Abutalebi, J.The neural basis of first and second language processing. Current Opinion in Neurobiology, 15 , 202 6.
98 Pimsleur, P., & Pimsleur, B. (1992). C'est la vie (5th ed.). Orlando: Harcourt Brace Jovanovich College Publishers. Pinto, D., & Rex, S. (2006). The acquisition of the spanish prepositions por and para in a classroom setting. His pania, 89 (3), 611 622. Roslim, N., & Mukundan, J. (2011). An overview of corpus linguistics studies on prepositions. English Language Teaching, 4 , 125+. Sanz, C., & VanPatten, B. (1998). On input processing, processing instruction, and the nature of replic ation tasks. Canadian Modern Language Review/La Revue Canadienne Des Langues Vivantes, 54 (2), 263 273. Schmidt, R. (1995). Consciousness and foreign language learning: A tutorial on the role of attention and awareness in learning. In R. Schmidt (Ed.), Atte ntion and awareness in foreign language learning (pp. 1 63). Honolulu, HI: National Foreign Language Resource Center. Schultz, J. M., & Tranvouez, M. . (2010). RÃ© seau: Communication, integration, intersections . Upper Saddle River, NJ: Prentice Hall. Ske han, P. (2001). A cognitive approach to language learning . Hong Kong: Oxford University Press. Swain, M. (1985). Communicative competence: Some roles of comprehensible input and comprehensible output in its development. In S. M. Gass, & C. G. Madden (Eds. ), Input in second language acquisiton (pp. 235 53). Rowley, MA: Newbury House. Tremblay, A. (2011). Proficiency assessment standards in second language acquisition research. Studies in Second Language Acquisition, 33 (03), 339. doi:10.1017/S02722631110000 15 Vandeloise, C. (1993). Les analyses de la prÃ©position dans: Faits linguistiques et effets mÃ©thodologiques. Lexique, 11 , 15 40. VanPatten, B. (2002). Processing instruction: An update. Language Learning, 52 , 755 803. VanPatten, B., & Cadierno, T. (1993 ). Explicit instruction and input processing. Studies in Second Language Acquisition, 15 (2), 225 243. VanPatten, B., Collopy, E., & Qualin, A. (2012). Explicit information and processing instruction with nominative and accusative case in russian as a secon d language: Just how important is explanation? Slavic and East European Journal, 56 (2), 256 276.
99 Vygotskii, L. S. (., Kozulin, A., & NetLibrary, I. (1986). Thought and language (Translation newly rev. and edited / by Alex Kozulin. ed.). Cambridge, Mass: MI T Press. Weber Fox, C. M., & Neville, H. J. (2001). Sensitive periods differentiate processing for open and closed class words: An ERP study in bilinguals. Journal of Speech, Language, and Hearing Research, 44 , 1338 53. Wong, W. (2005). Input enhancement: From theory and research to the classroom . New York: McGraw Hill.
100 BIOGRAPHICAL SKETCH Ann Healy received her Bachelor of Arts degree in French from the University of Louisville in 2009. For her academic performance during this degree, she received the award for Outstanding Graduate from the College of Arts and Sciences. She received her Master of Arts degree in French and Francophone Studies from the University of Louisville in 2012. That same year she joined the Languages, Literatures, and Cultures de partment at the University of Florida with the intention of completing a Master of Arts in French and Francophone Studies, with a concentration in French Linguistics, in August, 2014 . During both graduate degrees, Ms. Healy served as a Graduate Teach ing Assistant. She has taught a total of 46 credit hours in various introductory courses offered by both universities. In 2013, she was designated a semi finalist in the Graduate Teaching Awards competition at the University of Florida, after having been nom inated by both her supervisor and department chair.