Fossilization versus Stabilization in L2 Acquisition: Evidence from Inflectional Morphology among Yorùbá L2 English

Fossilization versus Stabilization in L2 Acquisition: Evidence from Inflectional Morphology among Yorùbá L2 English Boluwaji Oshodi ABSTRACT: Fossilization describes the end state of an L2 learner’s grammar, usually adeviant form of the target language. Since L2 learners tend to develop different interlanguage grammars during acquisition, fossilization appears complex and problematic. Thus, the term stabilization, which represents the L2 learner’s interlanguage at aparticular point, was suggested. This study investigates the two concepts to determine the point of their occurrence in L2 learners’ interlanguage. Two sets of data (oral production and composition) collected at an interval of seven years from two groups of Yorùbá native speakers whose L2 is English (Beginner and Advanced) with afocus on inflectional morphology revealed that while some errors disappeared from the interlanguage of learners in both groups, they still committed similar types of earlier errors despite the seven-year interval but at alesser rate. This confirmed the existence of different levels of error, i.e., stabilization. The results revealed that after the interval, inflectional errors across both groups were not static (fossilized) but changed (stabilized) due to different factors. This confirms that, while the stabilization of L2 features can be identified and established, it is usually difficult to identify fossilized features. KEY WORDS: fossilisation, stabilization, L2, inflectional morphology, interlanguage, English, Yorùbá 1. INTRODUCTION Based on previous studies on second language (henceforth L2) acquisition (henceforth SLA), it has been established that L2 learners display different types of proficiency at different levels of acquisition, which in reality reflects different types of interlanguage grammar. According to White (2003), L2 learners’ grammar tends to change (in most cases improve) as their level of exposure and knowledge of the L2 increases. However, despite endless opportunities to improve, it was observed that L2 learners usually retain deviant structures and forms because such deviant forms continue to reappear in L2 performance long after they are thought to have been overcome. This implies that L2 learners’ grammar, which is aimed at anative-like form, exhibits deviations in the process of acquisition. Some of these deviations manifest across all groups of L2 learners irrespective of their level of competence. They occur among those at the non-native-like and near-native-like levels and even among those with anative-like level, which is very close to the native speakers’ level. The major question which has received attention is why these deviant forms still appear and at what point in the process of learning do these deviant forms remain indelible in L2 performance. Selinker (1972) proposes the concept of fossilization that BOLUwAJI OSHODI 43 he claims is responsible for the reappearance of the deviant forms (errors) and also attempts to indicate the particular point at which such errors become permanent in L2 performance. However, Long (2005) observes that fossilization appears too broad to account for the issue. He argues that when L2 errors appear to stabilize at some point it does not indicate fossilization because such errors are not permanent since they tend to be corrected or even disappear in L2 performance. Therefore, he suggests the concept of stabilization, which is an offshoot of fossilization, to account for the phenomenon. Since L2 learners exhibit different types of errors, this longitudinal study will attempt to establish if there is any point at which L2 learners appear to have attained the level of permanence where any form of improvement is assumed to be impossible or whether they tend to correct their errors over time with age and more exposure to the target L2. The goal is to establish whether L2 learners stabilize or fossilize. The focus is on morphological variability, which concerns inflectional errors among L2 learners. 2. MORPHOLOGICAL VARIABILITY AND INFLECTIONAL ERRORS According to Dulay and Burt (1974), Beck (1998), White (2003), Haznedar (2007), McCarthy (2008), Lardiere (2009, 2011), Santoro (2012), and Oshodi (2014), the issue of tense and agreement has remained one of the most significant discussions in SLA studies. Tense and agreement are the areas in which L2 learners exhibit obvious and persistent errors in their interlanguage grammars, and this applies to both children and adults, including those at advanced or native-like levels. The errors basically involve the inconsistent use of target-like morphology, i.e., inflectional morphemes, by L2 learners, which is better known as morphological variability in SLA circles. From earlier literature on L2, e.g., Dulay and Burt (1974), Bley-Vroman (1989), Lardiere (1998a), to recent studies, e.g., Santoro (2012), Oshodi (2014), Foote (2017), Gwilliams and Marantz (2018), Bosch et al. (2019) and Ciaccio et al. (2020), morphological variability has been defined as the “inability” of L2 learners to accurately and consistently use inflectional morphology. Inaccurate and inconsistent use here could mean non-use (omission), wrong use, as well as over-use of the morphemes involved. In L2 English, which is our focus in this study, it involves the variable use of inflectional morphemes like 3rd person singular -s, past tense -ed (regular and irregular), plural markers (-s, -es, -ies, -en) and free morphemes, such as determiners (a, the, Ø), demonstratives (this, those, these), auxiliaries (is, are, has, have) and gender markers (he, she, him, her, his, her) in obligatory contexts (i.e., contexts in which native speakers would naturally use them). Based on the large body of literature that has investigated morphological variability among L2 learners, the problem is confined to two syntactic domains: verbal and nominal. Morphological variability has received tremendous attention, particularly the question of what constitutes the source or cause of L2 inflectional errors. It has to be explicitly stated here that the goal of this study is not to examine the source or cause of morphological variability, but to examine the concepts of fossilization and stabilization, with afocus on errors in inflectional morphology, to establish the point 44 STUDIE ZAPLIKOVANÉ LINGVISTIKY 1/2022 of their occurrence among L2 learners, using data collected from Beginner and Advanced Yorùbá native speakers who are L2 learners of English. 3. FOSSILIZATION AND STABILIZATION Selinker (1972, p.209) defines fossilization as “amechanism which is assumed to also exist in the latent psychological structure (…). Fossilizable linguistic phenomena are linguistic items, rules, and subsystems which speakers of aparticular native language will tend to keep in their interlanguage relative to aparticular target language, no matter what the age of the learner or the amount of explanation and instruction received in the target language”. Selinker (ibid, p.215) observes that “acrucial fact which any adequate theory of second language learning will have to explain is the regular reappearance or re-emergence in interlanguage productive performance of linguistic structures which were thought to have been eradicated. This behavioural reappearance is what has led me to postulate the reality of fossilization and interlanguage grammars”. This process of reappearance of non-native-like structures believed to have been eradicated is what Selinker (ibid.) describes as ‘backsliding’ to the permanent non-native-like state and terms fossilization. Selinker (ibid.) assumes that these non-native-like reoccurring structures have become permanent in the L2 learners’ interlanguage, hence his belief that they have become fossilized. According to Long (2005), stabilization can be defined as amechanism that underlies surface linguistic material which speakers of aparticular L1 tend to keep in their interlanguage at aparticular point in time relative to aparticular target language, no matter the age of the learner or amount of explanation and instruction he/ she receives in the target language, but which has the tendency to change over time based on several linguistic and social factors, which include the individual L2 learner involved, length of exposure to the L2, context, task, age, etc. This is somehow different from fossilization. Bley-Vroman (1989) opines that the major and significant difference between stabilization and fossilization is permanence. While stabilization assumes changes in the interlanguage over time occasioned by different factors, particularly more exposure to the target L2, fossilization assumes apermanent deficit of the L2 interlanguage. Selinker and Lakshmanan (1992) observe that stabilization is the first sign of (putative) fossilization. According to Long (2005), this makes it difficult to determine the exact point or period where the L2 stabilized interlanguage is fossilized. This makes stabilization testable, while fossilization is abit more difficult to ascertain. Both fossilization and stabilization are crucial concepts in SLA because they emphasize the role of L1 in the development of interlanguage grammars in L2 acquisition. For example, if the Yorùbá advanced learners examined in this study still commit similar errors of morphological variability in their interlanguage with the same frequency rate after seven years, then we can assume that it is acase of fossilization. However, if they improve or commit similar errors at alesser rate in the second instance of data collection, it may be considered acase of stabilization. Long (1997, p.491) observes that “since no study has shown fossilization in L2 children it BOLUwAJI OSHODI 45 is aconcept which can only occur in adult L2 learners”. Selinker (1992) expatiates on fossilization and sees it as asituation where L2 learners cannot progress beyond aparticular stage. There are serious issues with Selinker’s assertion. For example, at what point does fossilization take place? Does it take place immediately when an L2 learner starts learning an L2 or at alater time in the process? Can L2 learners still improve after fossilization has taken place? If yes, how can fossilization be established? This study shall attempt to provide answers to these questions. 4. WHAT IS INTERLANGUAGE GRAMMAR? The definitions of fossilization and stabilization often include the phrase “interlanguage grammar”. According to Selinker (1972), interlanguage refers to intermediate states or intermediate grammars of an L2 learner’s language as it moves towards the target language. It is aproduct of acreative process driven by inner forces and interaction, influenced by the native language and input from the target language. The “interlanguage hypothesis” sees the L2 learner’s language as an independent and variable system, which contains elements of the first and second languages as well as its own distinctive ones (Selinker, 1992). Consequently, the errors committed may not be traceable to the L1 and can also deviate from the normal L1 acquisition pattern. However, there are always obvious structural similarities between the two languages (i.e., the L1 and the target L2), and the resultant transfers and interferences are still highly prominent. The interlanguage grammar has several distinctive characteristics, the most prominent one being regular and frequent changes. Also, interlanguage is governed by some innate rules as well as the structure of the L1. Saville-Troike (2006) observes that interlanguage reduces complex grammaticality in form and it is also used for asmaller range of communicative needs. White (2000a) identifies three main levels of interlanguage: initial state, intermediate state, and end state. Interlanguage grammar also varies depending on several linguistic, cognitive, or even social factors among L2 learners with the same native language and asimilar background, experience and exposure to the L2 (White, 2000b). 5. PREVIOUS STUDIES ON FOSSILIZATION AND STABILIZATION Alarge number of studies have examined the concept of fossilization among different groups of L2 learners. Long (2005) observed that virtually all the studies on fossilization have come up with controversial findings which end up complicating the issue rather than resolving it. Previous studies on fossilization include Mukkatash (1986), Thep-Ackrapong (1990), Lennon (1991a, 1991b), Washburn (1992, 1994), Lin (1995), Lin and Hedgcock (1996). All these studies concluded that based on the persistent errors committed by the L2 learners examined, they have all reached astage where their interlanguage can no longer improve which by implication is asign of fossilization. Washburn (1992) and Lin (1995) both used the term fossilization for persistent L2 errors because they 46 STUDIE ZAPLIKOVANÉ LINGVISTIKY 1/2022 believe the learners have lived in the L2 environment long enough for such errors to still manifest in their interlanguage. However, serious issues have been raised concerning most of the findings of the studies that identified fossilization. The issues include the following points: (i) Results are based on assumptions because neither fossilization nor stabilization was actually demonstrated. (ii) There are problems with the types of subjects used. In most cases, inappropriate subjects were selected for such studies. (iii) The findings of these studies are based on insufficient data. (iv)Most analyses used in the studies are inadequate. Han (2009) proposes the Selective Fossilization Hypothesis (SFH), an analytic model that attempts to account for both the acquisitional and fossilizable prospect of linguistic features. The unique thing about this model is the recognition of the fact that not all L2 features are fossilizable and some features (e.g., morphosyntactic features) that have been the focus of most fossilization studies are more prone to fossilization than others. The major goal of SFH is to identify fossilizable features by investigating the likely factors that trigger fossilization in the acquisition process. According to Han’s (ibid.) Selective Fossilization Hypothesis, it is the interaction of first language markedness and the strength of the second language input taking place in the learner’s mind that determines how acquirable or fossilizable certain linguistic features will be. In summary, Han (ibid.) identified aselective type of stabilization in L2 interlanguage. Some studies have also examined and justified stabilization, which has been suggested by Long (1997) as abetter alternative to fossilization. They include Long (1997, 2005), Han (1998, 2000), and Lardiere (1998a, 1998b, 2000). These are all longitudinal studies in which L2 learners examined over atento twenty-year period showed evidence of improvement in the target L2 over time, even when they still committed the same type of errors. For example, Lardiere (1998a) examined Patty, an adult female native Chinese speaker who moved to the United States at the age of 22. The examination of her L2 (English) began after she had spent ten years learning it. This examination, which eventually lasted for nearly twenty years, revealed some interesting facts. The findings showed that although some errors still surfaced even after such along exposure to English, Patty exhibited signs of improvement and some errors disappeared from her interlanguage. Patty’s L2 interlanguage appeared to stabilize rather than fossilize because she would sometimes correct inflectional errors which she had committed earlier. The issue of what constitutes fossilization and stabilization remains contentious and more studies are needed to establish the point of their emergence. This study intends to contribute to this debate by using verifiable results, the right subjects, sufficient data, and valid analyses. 6. MORPHOLOGICAL VARIABILITY IN YORÙBÁ AND ENGLISH Yorùbá is aNiger-Congo language majorly spoken in the Southwestern part of Nigeria and English is the official language in Nigeria. Yorùbá native speakers in Nigeria speak English as asecond language. Unlike in English, where verbs and nouns are inflected for tense and number, verbs and nouns are never inflected in Yorùbá. For BOLUwAJI OSHODI 47 example, while the past tense form of the words go, come, eat, sell and kill are went, came, ate, sold and killed respectively, the Yorùbá versions l ‘go’, wá ‘come’, j ‘eat’, tà ‘sell’ and pa ‘kill’ remain the same when used in the past. In English, nouns are inflected for number. For example, one house becomes two houses in plural, one child becomes two children. In Yorùbá, however, nouns are not inflected. Ilé kan ‘one house’ remains ilé méjì ‘two houses’, m kan ‘one child’ remains m méjì ‘two children’, i.e., ilé ‘house’ and m ‘child’ maintain their form when used in the plural. Also, while English marks pronouns for gender, Yorùbá does not. In Yorùbá, the pronoun ó is used for male, female, and neuter while in English, he, him, his are used for male, she and her for female, and it for neuter. These examples clearly show that inflection is afeature of English that is not attested in Yorùbá. Several studies such as Afolayan (1968), Bamgbose (1982), Bamiro (1991), Babalola and Akande (2002), Akande (2003), and Alo and Mesthrie (2008) have examined and confirmed that Yorùbá native speakers whose L2 is English usually have problems with some aspects of the English language and the most prominent one is inflectional morphology, where errors are obvious and persistent. 7. DATA COLLECTION AND METHODOLOGY Morphological variability has focused mainly on two domains: verbal and nominal. Beck (1998), White (2003), and Lardiere (2005) examined morphological variability in the verbal domain focusing on finiteness, tense, and agreement (features which are expressed by -s), past tense markers (regular and irregular), auxiliaries (is, are, was, were, has, have), the occurrence of finite verbs in non-finite contexts and past tense verbs in present contexts. In the nominal domain, the focus has been on articles (definite, indefinite and null), demonstratives, plural marking, gender (masculine used in feminine contexts and feminine used in masculine contexts). These features shall be examined in this study. The features were coded in terms of error of accuracy versus inaccuracy, i.e., correct versus incorrect use of the inflectional features with afocus on the incorrect forms. The results were divided into the following categories: third person singular-s (3psg-s), past tense (regular and irregular), auxiliaries (is, are, has, have), finite verbs in non-finite contexts, articles (definite, indefinite, and null), demonstratives, plural marking, gender (masculine in feminine contexts and feminine in masculine contexts). The overall number of correct and incorrect usages of the inflections were calculated and will be presented as error count and accuracy rates in percentage. The data were collected in two instances. The first set was collected in 2012 while the second set was collected in 2019, which means there is aseven-year interval between both instances of data collection. 7.1 PARTIcIPANTS Sixteen participants, all native speakers of Yorùbá, took part in this study. They were divided into two groups, Beginner Group and Advanced Group, based on their age and 48 STUDIE ZAPLIKOVANÉ LINGVISTIKY 1/2022 length of exposure to the target L2. The first group comprised primary school pupils between eight and nine years of age who were still within the critical period (under the age of 10), and the second group was made up of adults between 40 and 50 years of age who were all secondary school teachers. The sixteen (16) subjects, eight (8) in the Beginner Group and eight (8) in the Advanced Group, were selected after considering factors such as the viability of subjects and the number of features to be examined. Having subjects who can provide valid data was crucial to the study. It was ascertained that the parents of all the subjects were native Yorùbá speakers. They all affirmed that their language of interaction at home (i.e., among family members) was Yorùbá. Their parents were confirmed to be artisans and petty traders who were not exposed to the English language formally. Also, it was verified that no member of the Advanced Group taught English language as asubject. During the first round of data collection, subjects in the Beginner Group were in their fifth year of primary school. Based on the information supplied by their parents, the researcher was able to gather some facts about their background in terms of their exposure to English. The parents affirmed that all the participants were never exposed to English before starting school, at least at home since they used (and still use) Yorùbá for communication at home. In the Nigerian primary school system, the fifth year is the period when children get their first formal exposure to the English language because pupils are taught in the mother tongue for the first three years and English is introduced in the fourth year alongside the mother tongue. There is one very important issue here that must be clarified. Considering the role of English as the official language in Nigeria, it may be difficult to determine the actual initial state of its acquisition because learners might pick up some English words through sources like the media (i.e., television and radio) long before they start the main process of formal learning. The subjects in the Advanced Group were all university graduates from different academic fields. They were all secondary school teachers in Yorùbá speaking towns during data collection. They had received both naturalistic and classroom exposure to English. They had had contacts with native speakers of English at one time or another during their secondary school and undergraduate days. Also, in their present employment, the medium of instruction is English. Based on the length of their exposure to the English language, their grammar was considered to be in an advanced state. They had passed two internationally recognized examinations: the General Certificate Examination (GCE O/L), which is the final year certificate examination taken in high school, and the Joint Matriculation Examination (JME), which is the university entrance examination in Nigeria, both conducted in English. Also, they had attended conferences and seminars conducted in English. Subjects in this group had had contact with the English language both formally and informally for over 37years. In the second round of data collection, some of the variables had changed. Members of the Beginner Group were now sixteen (16) years old and were all in senior secondary school 3, which is the certificate class. By implication, the length of their exposure to English had increased. Fortunately, the eight participants in BOLUwAJI OSHODI 49 this group were in the same secondary school. This made the second round of data collection easier. For the Advanced Group, some variables had also changed. Apart from the increase in age, two had become school principals and two had become vice-principals while the remaining four had become senior teachers. The length of their exposure to English had also increased. Only two members of the Advanced Group were in the same school during the second round of data collection. For the remaining six, the researcher had to conduct the two tests individually in their new schools. 7.2 PROcEDURE AND TASKS Two types of tasks were administered to the subjects: spontaneous oral production and written composition tasks. For the oral production task, subjects were tested in naturalistic contexts through interviews and discussions administered by the researcher. The subjects were allowed to discuss any topic and also ask questions about any topic that was of interest to them. This allowed us to elicit naturalistic data as much as possible. While the subjects in the Advanced Group were interviewed at home (individually) without any issues, subjects in the Beginner Group were reluctant to be interviewed at home. As aresult of this, the interviews and discussions were conducted in their school. Topics discussed varied according to the groups. For example, while the Beginner Group discussed topics such as The best movie Iever saw, My first day at school, My best friend, My room, My parents, and Two stories Ilike so much, those in the Advanced Group preferred discussions related to politics, salary, and current affairs. All the subjects in the two groups were interviewed individually four times within the period of one month and each interview lasted between 30 and 45 minutes. These interviews were recorded to make subsequent analysis possible. Following the method adopted by McCarthy (2008), utterances that were followed by self-expression were excluded; however, the final corrected forms were included. Self-repetitions were also excluded. In the written composition task, the subjects were instructed to pick atopic and write on it. The composition test was administered exactly amonth after administering the oral production test. Here, subjects in the Beginner Group wrote on one of these topics: The food Ilike best, My best friend, Agood movie Isaw, My school and My parents. These topics were chosen for members of this group because during the oral production tasks, the subjects showed tremendous interest in them. By implication, it means they tended to express themselves freely and produce natural sentences while writing on them. Subjects in the Advanced Group all agreed and wrote on the topic The Nigerian political situation. The written compositions were untimed and they were conducted under the supervision of the researcher. The fact that the tasks were untimed gave the subjects the opportunity to go over their answers and to have enough time to complete the tasks at their own pace and convenience. Another interesting fact is that subjects in both groups wrote on the same topics in both instances of data collection, i.e., in 2012 and 2019. 50 STUDIE ZAPLIKOVANÉ LINGVISTIKY 1/2022 8. RESULTS The results presented here focus on morphological variability of verbal and nominal inflections among the two groups of Yorùbá L2 English subjects based on elicitations of natural speech and writing. Results of two sets of data are presented. The first one reports the results of the tasks from the first year of data collection and the second one presents the results from the seventh year. Both results will be compared to see if there were changes between the 1stand the 7th-year results. This will help to establish if there were any changes in the rate and pattern of errors within the two groups as well as between them. Below are examples of inflectional errors in the interlanguage of the subjects (Beginner and Advanced Groups). The results are presented in two parts: those for data collected in the first year, i.e., in 2012, and those collected seven years after in 2019. They are presented under oral production and written composition. The focus is on the incorrect use of the target inflections in obligatory contexts. Results for auxiliaries are presented separately because they are free morphemes and are also inflected for tense. 8.1 ERRORS IN ORAL PRODUcTION (FIRST INSTANcE OF DATA cOLLEcTION) Examples of errors from the oral production tasks for both the Beginner and the Advanced Groups covering verbal and nominal inflections are presented below: 3rd person singular -s(agreement), past tense markers (regular and irregular) a. My mummy beat me anytime Idon’t read my book (beats) [sub2 BG] b. Ilikes to wear the shoe that my mummy bought for me (like) [sub4 BG] c. My mummy cook food for me which Ibrings to school (cooks, bring) [sub6 BG] d. Iknows the boy very well he does not come to school regularly (know) [sub3 AG] e. He fight alot yet he knows nothing (fights) [sub4 AG] f. Why does he likes to come late to school all the time? (like) [sub5 AG] Auxiliaries a. My daddy have bought acar this year (has) [sub1 BG] b. Everyone in my class have paid their school fees (has) [sub5 BG] c. Some people has been teaching for years yet with no experience (have) [sub6 AG] d. The boy join agroup that are robbing people in the town (is) [sub7 AG] Finite verb in non-finite contexts a. Yesterday my mother asked me to fetched water for her (fetch) [sub6 BG] b. When the man wanted to slaughtered the goat (slaughter) [sub4 BG] BOLUwAJI OSHODI 57 52 in contrast to 192 recorded in the first round for oral production and 21 in the second round in contrast to 40 recorded in the first round for written composition. Also, in Tables 3 and 4 (oral production and written composition: Advanced Group), the total number of incorrect uses of the inflections in the second round of data collection was 71 in contrast to 174 recorded in the first round for oral production and 10 in the second round as in contrast to 21 recorded in the first round for written composition. 8.4.2 cOMPARISON OF INcORREcT USE OF INFLEcTIONS: BEGINNER GROUP AND ADVANcED GROUP, 1ST ROUND The results of the comparison of the incidence of incorrect use of inflections between the Beginner and the Advanced Groups in oral production and written composition (1st round) are presented below in Tables 5 and 6. In Tables 5 and 6, two facts were also revealed. Firstly, in both tasks, i.e., oral production and written composition, the subjects in the Advanced Group had ahigher number of uses of the inflections than those in the Beginner Group. For example, in Table 5 (oral production), subjects in the Beginner Group used the target inflections 1887 times while those in the Advanced Group used the target inflections 3767 times. For written composition tasks, as revealed in Table 6, subjects in the Beginner Group used the target inflections 965 times while the subjects in the Advanced Group used the target inflections 1417 times. In both tasks, the subjects in the Advanced Group used the target inflections more than the subjects in the Beginner Group. Secondly, along with having ahigher number of occurrences of the target inflections, the subjects in the Advanced Group had alower number of errors than their counterparts in the Beginner Group. In Table 5, for oral production tasks, though the subjects in both groups had the same number of overall incorrect uses (which was 163 times), the overall percentage of errors was 4% for those in the Advanced Group and 10% for those in the Beginner Group. In Table 6, for the written composition tasks, the subjects in the Advanced Group had atotal of 21 instances of incorrect use in contrast to 40 incorrect uses exhibited by the subjects in the Beginner Group. 8.4.3 cOMPARISON OF INcORREcT USE OF INFLEcTIONS: BEGINNER GROUP AND ADVANcED GROUP, 2ND ROUND The results of the comparison of the incidence of incorrect use of inflections between the Beginner and the Advanced Groups in oral production and written composition (2nd round) are presented below in Tables 7 and 8. In Tables 7 and 8, two facts were revealed across the board. Just as in the first instance of data collection, in both tasks, i.e., oral production and written composition, the subjects in the Advanced Group had ahigher number of uses of the inflections than those in the Beginner Group. For example, in Table 7 (oral production), the subjects in the Beginner Group used the target inflections 3099 times while those in the Advanced Group used the target inflections 6369 times. Also, in Table 8 (written composition), subjects in the Beginner Group used the target inflections 1525 times while those in the Advanced Group used the target inflections 3178 times. 58 STUDIE ZAPLIKOVANÉ LINGVISTIKY 1/2022 1st Round 1st Round Beginner Group Advanced Group Correct Incorrect Incorrect % Correct Incorrect Incorrect % 3psg-3 92 23 20 151 26 15 Past tense marker (regular) 81 5 6 121 9 7 Past tense marker (irregular) 85 13 13 128 8 6 Auxiliaries (is, are, has, have) 201 11 5 521 25 6 Finite verbs in non-finite contexts 192 21 10 302 11 4 Articles (definite) 229 11 5 312 4 1 Articles (indefinite) 108 4 4 288 18 0.4 Articles (null) 33 7 18 101 16 3 Demonstratives 35 18 15 52 10 7 Gender (masculine used in feminine contexts) 311 32 9 874 19 2 Gender (feminine used in masculine contexts) 112 18 14 223 17 7 Gender feminine 93 0 0 312 0 0 Total 1695 163 10 3604 163 4 Table 5:Comparison of the incidence of incorrect use of inflections between the Beginner Group and the Advanced Group: Oral production (1st round). 1st Round 1st Round Beginner Group Advanced Group Correct Incorrect Incorrect % Correct Incorrect Incorrect % 3psg-3 68 8 16 102 3 3 Past tense marker (regular) 89 3 6 121 2 1 Past tense marker (irregular) 76 5 6 119 3 2.5 Auxiliaries (is, are, has, have) 112 4 4.3 203 1 1.5 Finite verbs in non-finite contexts 96 2 7 101 1 0.3 Articles (definite) 101 3 4.5 111 2 2 Articles (indefinite) 88 3 5 102 0 0 Articles (null) 33 2 11 49 2 7 Demonstratives 41 3 7 63 2 7 Gender (masculine used in feminine contexts) 74 4 9 201 3 2 Gender (feminine used in masculine contexts) 82 3 9 115 2 3 Gender feminine 65 0 0 109 0 0 Total 925 40 4 1396 21 2 Table 6:Comparison of the incidence of incorrect use of inflections between the Beginner Group and the Advanced Group: Written composition (1st round). BOLUwAJI OSHODI 59 2nd Round 2nd Round Beginner Group Advanced Group Correct Incorrect Incorrect % Correct Incorrect Incorrect % 3psg-3 246 16 6 305 10 2 Past tense marker (regular) 201 5 5 225 0 0 Past tense marker (irregular) 191 0 0 291 0 0 Auxiliaries (is, are, has, have) 481 0 0 899 0 0 Finite verbs in non-finite contexts 286 0 0 611 0 0 Articles (definite) 374 0 0 518 0 0 Articles (indefinite) 222 3 1 402 8 0 Articles (null) 51 2 4 224 9 0.4 Demonstratives 71 9 4 153 6 1 Gender (masculine used in feminine contexts) 335 7 2 1134 7 0.4 Gender (feminine used in masculine contexts) 214 8 4 493 0 0 Gender feminine 189 0 0 607 0 0 Total 3049 50 1.7 6329 40 0.6 Table 7:Comparison of the incidence of incorrect use of inflections between the Beginner Group and the Advanced Group: Oral production (2nd round). 2nd Round 2nd Round Beginner Group Advanced Group Correct Incorrect Incorrect % Correct Incorrect Incorrect % 3psg-3 120 3 6 412 3 0.7 Past tense marker (regular) 102 3 3 310 0 0 Past tense marker (irregular) 107 0 0 152 0 0 Auxiliaries (is, are, has, have) 270 0 0 356 0 0 Finite verbs in non-finite contexts 101 0 0 403 0 0 Articles (definite) 103 0 0 424 0 0 Articles (indefinite) 121 3 2 177 0 0 Articles (null) 58 2 3 114 3 1 Demonstratives 62 0 0 105 0 0 Gender (masculine used in feminine contexts) 211 6 2 354 4 0.5 Gender (feminine used in masculine contexts) 103 0 0 214 0 0 Gender feminine 150 0 0 147 0 0 Total 1508 17 1.1 3168 10 0.3 Table 8:Comparison of the incidence of incorrect use of inflections between the Beginner Group and the Advanced Group: Written composition (2nd round). 60 STUDIE ZAPLIKOVANÉ LINGVISTIKY 1/2022 Also, along with having ahigher use of the target inflections, the subjects in the Advanced Group still had alower number of errors than their counterparts in the Beginner Group. In Table 7, for the oral production tasks, subjects in the Advanced Group had atotal of 40 incorrect uses in contrast to 50 incorrect uses by those in the Beginner Group. While in Table 8, for the written composition tasks, the subjects in the Advanced Group had atotal of 10 instances of incorrect use in contrast to 17 incorrect uses exhibited by the subjects in the Beginner Group. 9. DISCUSSION The results presented in the previous section revealed very interesting findings across verbal and nominal inflections. As shown in Tables 1–4, in the first round of data collection, both in the oral production and in the written composition tasks, apart from the error of feminine gender in the masculine context, which was not committed, every subject in both the Beginner and the Advanced Groups committed errors in the use of other inflections tested, i.e., 3rd person singular -s, past tense (regular and irregular), auxiliaries (is, are, has, have), finite verbs in non-finite contexts, articles (definite, indefinite, null), demonstratives, plural marking, and gender (masculine used in feminine contexts). However, in the second round of data collection, three important facts were discovered. Firstly, some errors disappeared in the interlanguage of the subjects in the two groups both in the oral production and the written composition tasks. For example, in the second round of data collection among the Beginner Group, errors involving past tense (irregular verbs), auxiliaries, finite verbs in non-finite contexts, and articles (definite) all disappeared from their interlanguage in both the oral production and the written composition tasks. For the Advanced Group, errors involving past tense (regular and irregular verbs), auxiliaries, finite verbs in non-finite contexts, articles (definite and indefinite), and gender (masculine used in feminine contexts) all disappeared from their interlanguage in the two tasks examined. Secondly, in contexts where the same type of error was committed by the subjects in both groups and both instances of data collection in the two tasks examined, the rate of errors in the second instance of data collection was far lower than in the first instance for all the inflections tested (see Tables 1–4). Thirdly, in both instances of data collection and the two tasks examined, the subjects in the Advanced Group had ahigher rate of use of all the features tested and at the same time had alower rate of errors than the subjects in the Beginner Group (see Tables 5–8). Another important observation was the disparity between the results of the oral production tasks and in the written composition tasks, specifically in the incorrect use of demonstratives (this and these). In the oral production tasks in both instances of data collection and in the written composition in the first round of data collection, there were instances of the use of this /’ðɪs/ instead of these /’ði:z/ for plural nouns (see examples in section 8.1, 8.2) by the subjects in both groups. However, such errors did not occur at all in the written composition task of all the subjects after BOLUwAJI OSHODI 61 the seven-year interval. The use of /i/ and /s/ for /i:/ and /z/ by the subjects is no doubt aphonological issue where the case of fossilization may be established. The long vowel /i:/ and the consonant /z/ in these /’ði:z/ are not attested in Yorùbá. The subjects substituted them for /i/ and /s/, which are the closest vowels to them in Yorùbá. This explains why they committed these errors in the oral production tasks but did not commit them in the written composition task. The fact that they were able to distinguish between this and these in the written composition task after the seven-year interval, where these was used only with plural subjects, makes the issue of whether /i:/ and /z/ have been fossilized in the subjects’ interlanguage somehow difficult to establish. From the discussion of the results above, it is obvious that there were changes in the pattern as well as the rate of errors in the interlanguage of the subjects in the two groups examined across the two tasks (oral and written). The first instance of data collection represented alevel at which inflectional errors were identified and established. This according to Long (2005) is asign of stabilization. The second instance of data collection, which was done seven years later, also represented another level. At this point, there were improvements in the type and rate of the earlier identified errors. Some errors disappeared completely from the interlanguage of the subjects. This is most likely due to more exposure to the target L2 as aresult of the seven-year interval. This also substantiates acase of stabilization. Furthermore, subjects in the Advanced Group showed more improvement in the second round of data collection in terms of performance. For example, past tense (regular verbs), articles (indefinite), and gender (masculine used in feminine contexts), which still manifested in the interlanguage of the subjects in the Beginner Group at the second instance of data collection, disappeared from the interlanguage of the subjects in the Advanced Group. However, some of the earlier errors still manifested after seven years, but this time at alower rate. What this implies is that, despite the seven-year interval, which created room for more exposure to the target L2, some inflectional errors still fluctuated between appearance and non-appearance, i.e., these features sometimes appeared, and in some cases failed to appear in the interlanguage of the subjects. According to the theory of fossilization, fossilized features are permanently deficient, i.e., they never appear in the interlanguage of L2 learners. However, the intermittent appearance of such features against their assumed permanent non-appearance, as evident in the results above, makes it difficult to ascertain whether or not such features are truly fossilized. 10. CONCLUSION The concepts of fossilization and stabilization among L2 learners remain highly controversial. Though Long (2005) claimed that stabilization is an offshoot of fossilization, the issue of permanence, where fossilized errors are considered permanent errors, makes fossilization ahighly controversial concept. The argument would be that if interlanguage changes, i.e., errors that manifested at aprevious level are corrected at alater level in the interlanguage of L2 learners, then it would 62 STUDIE ZAPLIKOVANÉ LINGVISTIKY 1/2022 be incorrect to assume that L2 features can fossilize, because fossilization is apermanent thing. As shown in this study, apart from exhibiting alower rate of errors, the Yorùbá L2 English subjects in both groups were able to correct some of the inflectional errors they committed after seven years. This appears to be possible due to more exposure to English over the seven-year period. The two instances of data collection represent different levels of stabilization in their interlanguage. The fact that the subjects used some inflectional features correctly in certain contexts but used them incorrectly in other contexts after the seven-year interval does not truly demonstrate fossilization. If the features are truly fossilized, there should not be any instances of correct use by the subjects. There is atendency suggesting that alittle more exposure to the target L2 may lead to the disappearance of such errors from the subjects’ interlanguage. In conclusion, while every L2 linguistic feature can stabilize at apoint, not all L2 features can fossilize. According to the Selective Fossilization Hypothesis (SFH), only certain linguistic features tend to fossilize. However, since L2 learners can also improve with the right input and more exposure to the L2, the exact point at which an L2 feature fossilizes is difficult to establish. REFERENCES Afolayan, A. (1968). The linguistic problems of Yorùbá learners and users of English. Unpublished Ph.D. Dissertation, University College London. Akande, A.T. (2003). Acquisition of the inflectional morphemes by Nigerian learners of English language. Nordic Journal of African Studies, 12(3), 310–326. Alo, M., & Mesthrie, R. (2008). Nigerian English: Morphology and syntax. In R.Mesthrie (Ed.), Varieties of English, Africa, South and Southeast Asia (pp. 323–339). Mouton de Gruyter. Babalola, E.T., & Akande, A.T. (2002). Some linguistics problems of Yorùbá learners of English in Nigeria. ES: Revista de Filologia Inglesa, 24, 245–257. Bamiro, E. (1991). Nigerian Englishes in Nigerian English literature. World Englishes, 10(1), 7–17. https://doi.org/10.1111/j.1467-971X.1991. tb00133.x Bamgbose, A. (1982). Languages in contact: Yoruba and English in Nigeria. Education and Development, 2(1), 329–341. Beck, M.-L. (1998). L2 acquisition and obligatory head movement: English speaking learners of German and the local impairment hypothesis. Studies in Second Language Acquisition, 20, 311–348. https://doi.org/10.1017/ S0272263198003027 Bley-Vroman, R. (1989). What is the logical problem of foreign language learning? In S.Gass & J.Schachter (Eds.), Linguistic Perspectives on Second Language Acquisition (pp.41–68). Cambridge University Press. Bosch, S., Veríssimo, J., & Clahsen, H. (2019). Inflectional morphology in bilingual language processing: An age‐of‐acquisition study. Language Acquisition, 26, 339–360. https://doi. org/10.1080/10489223.2019.1570204 Ciaccio, L.A., Kgolo, N., & Clahsen, H. (2020). Morphological decomposition in Bantu: Amasked priming study on Setswana prefixation. Language, Cognition and Neuroscience, 35(10), 1–15. https://doi.org/10.10 80/23273798.2020.1722847 Dulay, H., & Burt, M. (1974). Natural sequences in child second language acquisition. Language Learning, 24, 37–53. https://doi. org/10.1111/j.1467-1770.1974.tb00234.x Foote, R. (2017). The storage and processing of morphologically complex words in L2 Spanish. Studies in Second Language Acquisition, BOLUwAJI OSHODI 63 39, 735–767. https://doi.org/10.1017/ S0272263115000376 Gwilliams, L., & Marantz, A. (2018). Morphological representations are extrapolated from morpho‐syntactic rules. Neuropsychologia, 114, 77– 87. https://doi. org/10.1016/j.neuropsychologia.2018.04.015 Han, Z.-H. (1998). Fossilization: An investigation into advanced L2 learning of atypologically distant language. Doctoral dissertation. Birkbeck College, University of London, Department of Applied Linguistics. Han, Z.-H. (2000). Fossilization in the State of the Art. Ms. Teachers College, New York. Han, Z.-H. (2009). Interlanguage and fossilization: Towards an analytic model. In V.Cook & L.Wei (Eds.), Contemporary Applied Linguistics. Vol. I: Language teaching and learning (pp. 137–162). Continuum. Haznedar, B. (2007). The acquisition of tenseaspect in child second language English. Second Language Research, 23, 383–417. https:// doi.org/10.1177/0267658307080330 Lardiere, D. (1998a). Case and tense in fossilized steady state grammar. Second Language Research, 14, 1–26. https://doi. org/10.1191/026765898674105303 Lardiere, D. (1998b). Dissociating syntax from morphology in adivergent L2 end-state grammar. Second Language Research, 14, 359–375. https://doi. org/10.1191/026765898672500216 Lardiere, D. (2000). Mapping features to forms in second language. In J.Archibald (Ed.), Second Language Acquisition and Linguistic Theory (pp. 102–29). Blackwell. Lardiere, D. (2005). On morphological competence. In L.Dekydtspotter, R.A.Sprouse & A.Liljestrand (Eds.), Proceedings of the 7th generative approaches to second language acquisition conference (GASLA 2004) (pp. 178–192). Cascadilla Proceedings Project. Lardiere, D. (2009). Some thoughts on the contrastive analysis of features in second language acquisition. Second Language Research, 25(2), 173–227. https://doi. org/10.1177/0267658308100283 Lardiere, D. (2011). Who is the interface hypothesis about? Linguistics Approaches to Bilingualism, 1(1), 48–53. https://doi. org/10.1075/lab.1.1.05lar Lennon, P. (1991a). Error and the very advanced learner. IRAL, 19(1), 30–44. Lennon, P. (1991b). Error elimination and error fossilization: Astudy of an advanced learner in the L2 community. ITL— International Journal of Applied Linguistics, 93/94, 129–51. https://doi.org/10.1075/itl.93-94.05len Lin, Y.-H. (1995). Un analysis empirico de la estabilizacion/fosilizacion: La incorporacion yla autocorreccion en unos sujetos chinos. Doctoral thesis, Departmento de Didactica de la lengua yla literature, Universidad de Barcelona. Lin, Y.-H., & Hedgcock, J. (1996). Negative feedback incorporation among highproficiency and low-proficiency Chinesespeaking learners of Spanish. Language Learning, 46(4), 567–611. https://doi. org/10.1111/j.1467-1770.1996.tb01353.x Long, M.H. (1997). Fossilization. Rigor motis in living linguistic systems? Plenary Address to the EORUSLA 97 conference. Universitat Pompeu Fbra, Barcelona, May 22–24. Long, M.H. (2005). Stabilization and fossilization. In C.J.Doughty & M.H.Long (Eds.), The Handbook of Second Language Acquisition (pp. 487–535). Blackwell. McCarthy, C. (2008). Morphological variability in the comprehension of agreement: An argument for representation over computation. Second Language Research, 24(4), 459–486. https://doi. org/10.1177/0267658308095737 Mukkatash, L. (1986). Persistence in fossilization. IRAL, 24(2), 187–203. Santoro, M. (2012). Morphological variability in interlanguage grammars: New evidence from the acquisition of gender and number in Italian determiner phrase and direct object pronouns. The Canadian Journal of Applied Linguistics, 15(1), 167–189. Saville-Troike, M. (2006). Introducing Second Language Acquisition. Cambridge University Press. 64 STUDIE ZAPLIKOVANÉ LINGVISTIKY 1/2022 Selinker, L. (1972). Interlanguage. International Review of Applied Linguistics, 10, 209–231. https://doi.org/10.1515/iral.1972.10.1-4.209 Selinker, L. (1992). Rediscovering Interlanguage. Longman. Selinker, L., & Lakshamanan, U. (1992). Language transfer and fossilization: The multiple effects principle. In S.M.Gass & L.Selinker (Eds.), Language Transfer in Language Learning (pp. 176–217). John Benjamins. Thep-Ackrapong, T. (1990) Fossilization: Acase study of practical and theoretical parameters. Unpublished doctoral dissertation, Illinois State University. Washburn, G. (1992). Fossilization in second language acquisition. AVygotskian analysis. Ph.D. dissertation, University of Pennsylvania. Washburn, G. (1994). Working in the ZPD: Fossilized and nonfossilized non-native speakers. In J.P.Lantolf & G.Appel (Eds.), Vygotskian Approaches to Second Language Research (pp. 69–81). Ablex. White, L. (2000a). SLA from initial to final state. In J.Archibald (Ed.), Second Language Acquisition and Linguistic Theory (pp. 130–155). Blackwell. White, L. (2000b). Universal Grammar in second language acquisition: The nature of interlanguage representation. In A.Juffs, A.T.Talplas, G.Mizera, & B.Burtt (Eds.), Proceedings of GASLA IV. Working Papers in Linguistics IV (pp. 3–14). University of Pittsburgh. White, L. (2003). Fossilization in steady state L2 grammars: Persistent problems with inflectional morphology. Bilingualism: Language and Cognition, 6, 129–141. https://doi. org/10.1017/S1366728903001081 Boluwaji Oshodi | Adekunle Ajasin University Akungba Akoko, Ondo State, Nigeria <[email protected]>