A Fossilized Language Pattern for the Singular First Person Subject Pronoun in the Saudi Context : I vs . I am

The drive that motivated this study was an English as a Foreign Language (EFL) teacher’s observation of a fossilized grammar pattern produced by students over an extended period of time. The fossilized pattern “I am go” frequently emerges in the EFL context of Saudi Arabia, for which a number of factors could be accountable, including overgeneralization of a grammar pattern, inadequate instruction, lack of negative and corrective feedback, being frequently exposed to peers’ production of the fossilized pattern, and so forth. 155 undergraduate students from a Saudi university responded to a multiple-choice question with three options. The findings revealed that only one third of the participants identified the correct singular first person subject pronoun in English “I”, which corresponded to (انا) in Arabic. Based on the results, pedagogical and methodological recommendations are made as to how the possibility of the emergence of the incorrect pattern in question can be reduced or minimized.


Introduction
The intention of this analysis is to shed light on a fossilized grammar pattern, commonly observed in L2 learners in the EFL context of Saudi Arabia.The study was conducted upon a teacher's observation of a fossilized grammar pattern very frequently produced by Saudi learners studying at a major university in Saudi Arabia.Saudi students commonly produce sentences such as "I am go" and "I am eat", which led the researcher to investigate why students consistently used the verb ASIAN TEFL, Vol. 3 No.1, 2018, www.asian-tefl.com"to be" in the form of "am" for the first person singular subject pronoun "I", followed by the infinitive form of the verb.The hypothesis was that students came to believe that in order to express the first person singular subject pronoun in English "I", they consistently produced the fossilized pattern "I am".In other words, to mean "I", students' interlanguage motivated the insertion of "am" beside the pronoun "I", thereby creating the fossilized pattern "I am go".
What makes the word "am" specifically more challenging for second language learners is that "am" can be used as either a content or function word, depending on the context.For instance in the sentence, "I am happy", the verb am is the main verb, whereas in the sentence "I am eating", am is used as an auxiliary verb.The verb "am" can also be regarded as a linking verb, in the sense that it does not describe any direct action taken or controlled by the subject.Unlike the majority of verbs, it is a verb which does not involve an action, but rather describes a state; all other forms of the verb to be have the same function.
All natural languages contain lexical morphemes (content words) such as nouns, verbs, adjectives and functional morphemes (function words) such as definite and indefinite articles, auxiliary verbs, modal verbs, prepositions (Stromswold, 1994).Content words are relatively more concrete and pictureable, or imageable compared to function words, which serve a syntactic role rather than a semantic one (Smith & Witten, 1993).Tyler, Moss, Galpin, and Voice (2002, p. 476) define imageability of a word as "the degree to which its referent can be perceived through the senses".For instance, the word "chair" is a word that is concrete and highly imageable word since its meaning is associated with many sensory properties (size, shape, etc.) whereas a word such as "belief" is low in imageability in that it does not invoke a concrete and imageable meaning.
The relevance of function words to this study is that function words are among structures which are relatively less perceptually salient, less concrete, and less pictureable (imageable), and are therefore more prone to fossilization compared to content words (Todeva, 1992).Chung and Pennebaker (2007), in addition, posit that function words among the hardest language units in the process of learning a second language.
This study, more specifically, aims to investigate why the pattern in question fossilizes, and attempts to identify the rationale behind learners' adopting an incorrect language pattern and integrating it into their interlanguage.The adoption of such a fossilized language pattern also runs the risk of interacting with other language patterns, complicating the acquisition process by triggering further fossilized patterns.If the cause of fossilized patterns can be assessed, then pedagogy can play a role in preventing fossilized patterns from emerging altogether rather than treating the error after it has become fossilized.In other words, instruction should be planned and implemented in a way so that the possibility of the emergence of fossilized patterns, at the very least, becomes slim.In this regard, it is of utmost importance for teachers to assess which grammar items need to be constantly emphasized and recycled until students' awareness is raised as to the correct versus incorrect language patterns.
To recap this introductory section, the aim of this study is twofold.Firstly, it attempts to identify the cause(s) of a fossilized pattern.Secondly, and more importantly, it tries to close a gap in the second language acquisition (SLA) literature.To the best of my knowledge, research on the fossilization of the function word "am" is non-existent.The Saudi context is no exception; no research has been undertaken so far to investigate the cause of the fossilized "I am go" pattern.Previous studies conducted with Turkish speakers (Snape & Kupisch, 2010;Abushihab, (2014) and Arabic speakers (Al-Jamal, 2017) have rather highlighted the difficulty of Turkish and Arabic speakers with the definite article the, in particular.Research, however, on the fossilization of the word "am" as an auxiliary, or main verb is rare, if not non-existent.This study will, in this regard, attempt to investigate the hypothesis whether function words are more prone to fossilization.Function words are discussed below in the section 2.3 in more detail.

Literature Review
There seems be to be a consensus in the second language acquisition (SLA) literature (Selinker, 1972;Bley-Vroman, 1989;Richard, 2008;Nushi, 2016) on the fact that fossilized error patterns resist correction not only in EFL contexts, but also in English as Second Language (ESL) contexts.In other words, learners living in English-language communities also adopt fossilized language patterns despite being constantly exposed to natural language input, which is also referred to as positive evidence.Positive evidence is defined as the set of well-formed sentences to which learners are exposed" (Gass, 1997, p. 36).For a second language learner, this shows which language structures are acceptable in the target language.Negative evidence, otherwise referred to as "negative feedback", informs the learner which language structure violates the correct target language structure.In a way, it is an explicit and direct way of telling the learner that the language pattern he1 has produced is incorrect, which is typically followed by a teacher's producing the acceptable language pattern, normally referred to as corrective feedback.Vigil and Oller (1976) posit that interactive feedback, defined as the combination of positive emotional feedback and negative cognitive feedback, is the best method of providing learners with feedback on their language errors.The former is intended to motivate the student and boost his confidence while the latter aims to provide the student with cognitive feedback that explicitly signals to the student that making changes to the pattern produced is a necessity.Hasbun (2007) indicates that some fossilized patterns are adopted in the early stages of acquisition while others require a longer period of time for repair, and some are persistent over time, which may defy pedagogic interventions resulting in permanent fossilization.The following section will provide some of the definitions of fossilization in the literature.

Definitions and Overview
Fossilization refers to the "often-observed loss of progress in the acquisition of a second language, following a period where learning occurred, despite regular ASIAN TEFL, Vol. 3 No.1, 2018, www.asian-tefl.comexposure to and interaction with the target language (L2) and regardless of any learner motivation to continue" (Banerjee, 2013, p. 40).If the learner has reached a stage in which a language structure does still not have the same form as the target language, the fossilization will manifest itself as an error (Han, 2004).Han defines fossilization as "the phenomenon of the non-progression of learning despite continuous exposure to input, adequate motivation to learn, and sufficient opportunity for practice" (p.213).Skehan defines fossilization (1998, p. 61) as "a misinterpretation of a rule by a learner which has not subsequently been eradicated by the accumulation of well-formed formulaic items".
If we are past the age of around 7-10 years, the acquisition of an L2, in marked contrast to the way we acquired our first language (L1), can turn out to be rather slow, laborious and, even in talented L2 learners, tends to stop short of native-like proficiency.This "stopping short" is referred to as fossilization (Selinker, 1972) or incompleteness (Schachter 1990, as cited in Towell & Hawkins, 1994, p. 2).Richards, Platt, Platt, (1992, p. 145) define 'fossilization' as a "process in which incorrect linguistic features become a permanent part of the way a person speaks or writes a language."Language properties that are most vulnerable to fossilization are vocabulary usage, grammar, and pronunciation.Richards et al. contend that a foreign accent is an obvious aspect of fossilized pronunciation.Selinker (1972, p. 215) postulates that fossilized linguistic structures "even when seemingly eradicated, are still somehow present in the brain, stored by a fossilization mechanism in an interlanguage".Azar (2002) posits that fossilization occurs when usage errors have become so embedded in L2 learners' interlanguage that it becomes habitual in their language production, which emerges as a consequence of learners' getting no corrective feedback.According to this scenario, an obvious consequence of fossilization is that learners create their own grammar patterns, which are constantly repeated and recycled since no input, or feedback is provided as to the fact that the pattern in question is incorrect.
As far as adult second language learning is concerned, in the SLA literature, there seems be to be a consensus on the fact that fossilization cannot be prevented.For instance, Selinker (1972, p. 229) contends that "learners tend to persist in making linguistic errors in their interlanguage no matter what the age of the learner is".Mitchell and Myles (1998, p. 13), in the same vein, assert that "the languagespecific mechanisms available to the young child simply cease to work for older learners", if not completely, then partially, and indicate that no amount of study and effort can recreate them.Ellis (1985), likewise, maintains that fossilized patterns occur in most languages and cannot be repaired by further instruction.
The preceding discussion that portrays fossilization as an irreversible and unrepairable process has been contested by Vigil and Oller (1976) who contend that fossilization may be overcome if the learner is given the necessary positive affective feedback, motivating the learner to further engage in communication, along with neutral or negative cognitive feedback, which is further elaborated on by Brown (1994, p. 218) who posits that learners should be encouraged to "try again", to restate, to reformulate or to draw a different hypothesis about a rule.Ferris, (2004) along the same line, argues that compared to those who receive no feedback, students who receive feedback on their errors will be more likely to self-correct them during revision.
As for what causes fossilization to occur, researchers have usually attributed this to the Critical Period Hypothesis, according to which a child is blessed with a brain which has an innate propensity for language acquisition, which atrophies as time passes (Lenneberg, 1972).DeKeyser (1993) argues that if no sufficient error treatment is provided, then fossilization is a naturally emerging consequence and is unavoidable.Han (2004, p.216), in addition, asserts that "no adult L2 learner would ever be able to pass for native in all contexts".According to Brown (1994), fossilization may be caused by the following two factors: a-borrowing patterns from the mother tongue, b-extending patterns from the target language, or over-generalizing a learner rule.Brown (1994, p. 222-223), on the other hand, categorizes fossilized errors into four groups: "addition, omission, substitution, and ordering at either the sentence or discourse level".Based on these categories, Brown posits that errors occur in different linguistic domains such as phonology, orthography, lexicon, grammar and discourse.Errors are also categorized based on whether they hinder communication (global) or not.If they do not hinder communication, they are referred to as "local" errors.
Yorio (1994( , cited in O'Riordan, 1999)), on the other hand, attributes language fossilization to an extended period of fluency without accuracy.Lightbown, Halter, White, & Horst (2002) posit that fossilization is more commonly observed in classes where students share the same native language.In such classes, since learners share the same native language and display similar interlanguage patterns, a lot of nonstandard input is generated, which tends to confirm their erroneous interlanguage hypotheses.
With regard to the fossilization categories above, Selinker (1974, p. 38) cites the example of an overgeneralized pattern; the past-tense morpheme -ed is added to go (an irregular verb), producing the fossilized pattern "goed".Selinker also cites an Indian speaker as another example who collocates 'drive' with 'bicycle'.The "goed" example does not usually occur in the Saudi context.Students in this context, however, do say "I sended a message", but not "I goed".It is obviously not easy to tell why certain fossilized structures fossilize, and others do not.That said, one can make predictions as to why certain fossilized patterns emerge the way they do.From my personal experience of teaching in the Saudi context for more than a decade, I can say that the word "sended" is adopted by students simply because learners in this context have difficulty memorizing the past and the past participle forms of irregular verbs.
As will be further discussed below, communication-first methodologies prioritize communication at the expense of accuracy.The word "sended" is, in all likelihood, will be understood by any native or fluent speaker of English; the message that needs to be communicated simply goes through.If communication is attained, students may adopt, a "why-bother" attitude since they are aware that they are comprehended by their interlocutors.This is only one example of a possible source of fossilized pattern as far as the Saudi context is concerned; the typological distance between Arabic and English is another convincing factor that might contribute to numerous fossilized patterns in addition to the "I am go" pattern.The discrepancy between the Arabic and English language structures is very likely to contribute to the formation of this particular pattern; whereas in English the sentence "I am a student" is expressed through four lexical items, the corresponding sentence in Arabic requires a two-word sentence only, which does not contain a verb.
The mode of instruction could also account for the fossilized pattern "I am go".
When we consider the fact that structures in the present tense occur much more frequently than either the past or the future tense, and considering the good possibility of students' having been exposed to grammar structures in the Simple Present Tense such as I am happy and I am a student a zillion times, it does make sense for students to believe that ‫انا‬ (the singular first person subject pronoun) in Arabic, corresponds to "I am".Hence, it may simply suffice for students to insert "go" beside "I am" thereby producing the I am go pattern.This fossilized pattern seems to justify Brown's category above, namely "overgeneralization", where a learner extends the use of a grammatical rule or linguistic item beyond its acceptable pattern, generally by making words or a structure follow a more regular pattern.In the example, "I am go", learners extend the "I am" pattern in the sentences "I am a student" and "I am happy" to "I am go" thereby overgeneralizing the "I am".In this overgeneralization process, learners create their own erroneous pattern and adopt the pattern "I am" to actually refer to "I".
When we put ourselves in the shoes of our students, on the other hand, we can see that it is easy for them to believe that in the sentences "I am happy" and "I am a student", "I am" corresponds to the "I" pronoun in Arabic ‫)انا(‬ for the very simple reason that the verb "to be" is not used in Arabic when a person wishes to say that I am a student.Both in spoken and written Arabic, all we hear is two words, ‫طالبٌ‬ ‫,أنا‬ which means I am a student.The indefinite article "a", to make things worse for students, does not exist in Arabic as well as in many other languages before a singular countable noun.Gass and Mackey (2002) add that some of the copular verbs which include different forms of the verb to be (am, is, are) do not exist in Arabic.Another important feature of Arabic is that, in contrast to the sentence structure in English, it is well possible to make sentences without a verb.

Function Words vs. Content Words
The words am and a are referred to as "function words", which are actually used to add grammaticality to the sentence.In other words, like many other function words in English (been, has, is, are, have, will), the words am and a do not substantially contribute to the meaning of the sentence, and to this end, they can simply be labeled as grammar words.As Foster and Ohta (2005) put it, morphosyntactic components of language do not substantially alter the meaning that interlocutors may wish to communicate.This is because missing, incorrect or unrecognized morphemes or problems with tense, case, or gender do not hinder communication; learners, therefore, can easily communicate meanings simply by employing lexis (content words) and contextual clues.This being the case, a knowledgeable teacher or linguist should be aware of the fact that function words have a great potential to undergo fossilization for students learning English as a target language (L2).It is obviously not so easy, as argued above, to predict why certain structures, or forms fossilize more easily than others.That said, based on the similarities and differences of structures of L1 and L2, we can definitely make informed predictions.

Methodological Causes of Fossilization
In the discussion above, it is acknowledged that methodological practices may well contribute to the adoption of fossilized language patterns.In theory, both meaningoriented (Communicative Language Teaching) and form-oriented (the Grammar-Translation) teaching methodologies may play a role in causing fossilized patterns.
There is a consensus in the SLA literature that meaning-focused teaching methodologies are regarded as the major cause of fossilization.For instance, Celce-Murcia; Dörnyei; and Thurrell (1997) contend that there is a perception that linguistic form emerges on its own as a result of learners' engaging in communicative activities.Bruton, (2002) also expresses his doubt about whether tasks as core activities resolve the fundamental questions of accurate oral production.Likewise, considerable number of SLA researchers believe that when second-language learners' attention is directed to solely meaning with no focus on form, linguistic accuracy suffers (Alderson & Steel, 1994;Harley & Swain, 1984;Lightbown & Halter, 1993;Lyster, 1987;DeKeyser, 1995;Ellis, 1993;Harley, 1989;Long, 1991;Robinson, 1996;Terrell, 1991;White, 1991, as cited in Renou, 2001).
Unless explicit instruction and feedback is provided on their errors, adult acquirers may not make progress in accuracy of linguistic forms (Doughty & Varela, 1998;Doughty & Williams, 1998;Ellis, 1998;James, 1998;Lightbown, 1998;Lyster & Ranta, 1997;Tomasello & Herron, 1989, as cited in Ferris, 2004).Zhang (1998), in addition, posits that meaning-focused methodologies such as Communicative Language Teaching can cause the learner to become a sophisticated user of his own pidginized communicative interlanguage, which constitutes a linguistic system itself.Lightbown and Spada (1999), along the same line, contend that in meaning-oriented methodologies, pedagogy does not substantially involve error correction and explicit instruction, which accounts for early fossilization of errors.Eskey (1983), furthermore, exclaims: "We cannot go on accepting inaccurate language simply because it communicates something that a clever native speaker can somehow understand" (p.322).A more student-centered approach comes from Ferris (2004), who posits that students do attend to and appreciate feedback on their errors, the lack of which may lead to anxiety or resentment, which could decrease motivation and lower confidence in their teachers (p.56).
Teachers who have been trained with a communicative orientation to teaching, may also adopt beliefs that may significantly affect their instruction and error correction practices, which is documented in the following quotation, taken from an interview with a teacher in the Australian context: "It doesn't really matter at first how many mistakes the students make as long as they are communicating" (Mangubhai, Marland, Dashwood, Son, 2004, p. 306).It is obvious that teachers who maintain such methodological views may place their students at a serious risk since errors that are not dealt with in the initial stages of acquisition, as discussed above, are very likely to fossilize and become permanent.

Research Design
As discussed above, the fossilized pattern "I am" which normally replaces its correct counterpart "I" is frequently produced by Saudi learners.It is important to point out that the hypothesis cited in the preceding sentence is based on an EFL teacher's personal observation only, and does not depend or follow up on any previous empirical investigation.
To test the hypothesis that the fossilized pattern in question indeed refers to the pronoun "I", the researcher (also the EFL teacher mentioned above), designed an instrument which contained one single question, based on multiple-choice question format.To attain reliability and to make sure that the learners understood what they were required to do to complete the task, a very critical linguistic component was added to the question.The precise Arabic equivalent of the subject pronoun "I" ‫)انا(‬ was provided to ensure that learners comprehended the question fully and knew what exactly they were required to do to complete the task.
Learners who participated in the study were intermediate-level students in terms of their English language proficiency, ranging from lower intermediate or upperintermediate proficiency levels.To this end, the question was so designed that all proficiency levels were catered for.The question was: Which of the following means ‫انا‬ in Arabic?a-I b-I am c-Both a and b Based on the experience of the teacher, it is the case that a good number of students believe that both I and I am correspond to the first singular person subject pronoun "I", the equivalent of ‫انا‬ in Arabic.For this reason, the both a and b option was integrated to the options.The researcher anticipated that the answer to the question above would determine whether "I am" replaces "I" and constitutes a fossilized pattern in students' interlanguage.
The question was delivered to the participants in two ways.The first was pen-andpaper based where students were personally approached and were asked to respond to the question in the presence of the researcher.As for the second method of delivery, Google Forms was used.In other words, the question was made online, where the participants could respond to the question at their convenience.
Once data collection was complete, informal interviews were conducted with three students; the students were interviewed based on their responses.A student was selected from each response category.A more clear distribution of the students interviewed was as follows: a-one student whose response was "I" b-one student whose response was "I am" c-one student whose response was both a and b, meaning that he thought both "I" and "I am" were correct.
To summarize the foregoing, to obtain data, the study used two research tools.The first one included the direct exposure of the students to the question, where they had to make a choice among the three options provided.The second tool was informal interviews, which aimed to understand why a student selected the specific option he thought was correct.This in turn would complement the findings obtained through their responses to the one-question research instrument.

The Findings
As the foregoing discussion suggests, in the EFL context of Saudi Arabia, learners tend to produce the "I am" pattern rather than the subject pronoun "I", which is evident in the fossilized pattern "I am go".The hypothesis was that learners adopt the "I am" pattern to actually refer to "I".This particular investigation was carried out to test whether this hypothesis would be justified.
Figure 1 and Table 1 below summarize the findings obtained from the study.The above graph and table above point to the following: Out of the 155 participants, 50 students knew the correct answer, suggesting that only 32% of the students had correctly acquired the singular first person subject pronoun "I" in English, corresponding to ‫انا(‬ ) in Arabic.25% of the students thought that only the "I am" pattern corresponded to the correct subject pronoun "I", while a significant number of students (67 out of 155 = 43%) felt that both "I" and "I am" corresponded to the singular first person subject pronoun in English.Reading the figures from a broader perspective, 68% (25% = I am + 43% = both) of the participants had acquired the fossilized pattern discussed above while about only one third of the participants (32%) were aware that only "I" corresponded to the correct subject pronoun.
As for the interviews, what students said justified the rationale for why students thought that "I am" corresponded to ‫انا‬ in Arabic.The following excerpts were taken from interviews with students.The names below are pseudonyms.
In particular, what Ali (the student who thought that the correct response was both I and I am) lends said strong support to the hypothesis; taking the sentence "I am a student" as the base, Ali thought that I am corresponded to I= ‫.انا‬ Below are the excerpts from the interviews.The implications of the findings obtained through the instrument and informal interviews are discussed in some detail below in the section titled Discussion.

Discussion
The aim of this study was to investigate why Saudi learners adopt the fossilized pattern "I am go" rather than "I go".More specifically, the rationale behind learners' insertion of the word "am" beside I was the focus of the inquiry.
The findings obtained from the study point to the fact that two thirds of the participants had adopted the fossilized pattern "I am go", while only 32 percent of the participants had acquired the correct subject pronoun.That only about one third of students had acquired the subject pronoun "I" correctly indeed deserves investigation as to what factors may have caused this fossilized pattern.
As discussed above, a variety of factors account for fossilized errors.Interview 3 with Ali clearly indicates that overgeneralization of a grammar pattern is what causes the pattern to occur.Before administering the study, as discussed above, I predicted that students' justification for "I am go" originated from the sentence "I am a student", where "I am" is extended, or carried over to the sentence "I am go".What Ali said in the interview lends support to my prediction.
As quoted above in Interview 3, when Ali says "I am Ali" to introduce himself, he thinks that the "I am" pattern represents the subject pronoun "I".If the context is different, however, Ali feels that only I could suffice to say ‫انا‬ in Arabic.To exemplify this, Ali quotes the question "did you do that?" to which the answer, he says, would be "Yes, I did".In the answer "Yes, I did", he emphasizes that only I would serve as the subject pronoun.Ali's misconception, in other words, indicates that depending on varying grammatical contexts, either "I" or "I am" would correspond to the subject pronoun I.
Interview 2, on the other hand, suggests that inadequate instruction, coupled with a lack of negative and corrective feedback, also seems to have contributed to the fossilized pattern.Mohammad's sentences "This is what I heard him (the teacher) say", "I did not ask him is it I or I am?", and "in English I…….I am, the same" clearly shows that instruction may have caused the fossilized pattern.It is evident that the students' error was possibly not addressed by the instructor.Higgs and Clifford's following quotation lends support to this hypothesis.
The terminal cases whose foreign-language background had included only an academic environment all came from language programs that either were taught by instructors who themselves had not attained grammatical mastery of the target language -and hence were unable to guide their students into correct usage -or by instructors who had chosen not to correct their students' mistakes for philosophical, methodological, or personal reasons.(1982, p. 68) It is indeed difficult to identify the exact reason for the fossilized pattern as more than one variable may play a role in students' producing the "I am" pattern, when actually the intended meaning corresponded simply to the subject pronoun I.
As Lightbown, Halter, White, & Horst (2002) point out, in monolingual classes, learners receive abundant deviant input from classmates, which confirms their erroneous assumptions, or hypotheses regarding the accuracy of the structures they produce.Lightbown et al. further argue that in contexts where learners' attention is focused primarily on understanding the meaning of a text or a conversation, learners fail to notice formal features of the target language.This suggests that in meaningoriented instruction, formal properties of language, i.e. function words such as articles and prepositions as well as morphosyntactic markers of verbs such as tense and aspect are often overlooked as if they were completely transparent.It is difficult to predict precisely what mode of instruction was delivered to students in the Saudi context as it is usually the case that instructors usually do switch between the meaning-oriented and form-oriented instruction.In other words, based on the needs of students, instructors may teach grammar (explicitly or implicitly) as well as engaging students in communicative activities such as pair work and group work activities.
The statistics of this study demonstrate that 68 % of the participants do not have a clear picture of what exactly represents the first person subject pronoun I.The learners who believed that I am (25%) represented the subject pronoun in addition to those who thought the subject pronoun could be expressed through either I or I am (43%) constituted two thirds of the participants.
These findings justify the literature and the hypotheses discussed in the Introduction.
Firstly, function words are perceptually less salient, less concrete, and less imageable (pictureable) words as hypothesized by Todeva (1992) and are far more challenging language structures for EFL learners to acquire compared to content words (Chung & Pennebaker, (2007).The above cited language features (salience, concreteness, and imageability) lead researchers to the conclusion that function words are also more prone to fossilization by comparison with content words (Todeva, 1992).Another important factor that impedes the learning process is the fact auxiliary verbs do not exist in Arabic (Al-Jamal, 2017).
The findings, for the Saudi context, compellingly points to an instant need for intervention; but how?The implications of this fossilized pattern and potential solutions will be discussed below.The section titled Conclusion will also involve a supplementary review of the literature based on the findings obtained from the study.In other words, a second literature review has been integrated in light of the findings.

Conclusion
There seems to be two opposing camps on grammar error correction.The first one is represented by researchers such as Truscott (1999) and Krashen (1982), who are advocates of no-grammar instruction/correction approach.Truscott argues that correction, especially as far as writing is concerned, should be abandoned since he feels that SLA research does not support error correction.In his view, error correction is not only ineffective, but also harmful.This, in reality, echoes the mindset in the 1960s.Braddock, Lloyd-Jones and Schoer (1963), for instance, claimed that the teaching of formal grammar has a harmful effect on the improvement of writing.Along the same vein, the Commission on English in the US (1965) reported that "traditional grammar teaching had a minor impact on writing, and that it did not contribute to the improvement in oral competence.Two decades later, the same hostile attitude (Schmidt, 1990) towards the teaching of grammar relapsed.The National Council of Teachers of English (NCTE) in the US declared that the teaching of grammar does not lead to any improvement in students' speaking and writing (1985).In addition to the preceding anti-grammar and anti-explicit error correction movement, there are also those researchers who tend to believe that error correction should be made only if the errors inhibited communication and stigmatized learners (Hendrickson, 1978).In other words, the teacher should not intervene unless there is a communication breakdown.
The second camp, on the other hand, is represented by numerous researchers, who advocate the provision of error feedback.Ferris (2004), for instance, posits that L2 learners value error correction and indeed desire feedback on their errors regardless of pedagogical focus and classroom setting.Ferris further argues that L2 learners feel that error correction is essential to their language development.
Another important factor that causes controversy over the error correction issue is the source of error correction.In other words, the question is: "who should provide error correction?" Hyland and Hyland (2006) explain that there are again two opposing views regarding the source of error correction.There are teachers, for instance, who reject direct explicit correction on students' errors on the grounds that students should be encouraged to be independent, and that they should attempt to self-correct.In essence, meaning-oriented methodologies promote self-correction, which is defined as "the technique in which students are engaged to correct their own errors".(Rief, 1990, cited in Sultana, 2009, p. 11).The opposite side of the coin, on the other hand, points to teachers who support the idea of intervening and providing feedback whenever students are in need.Hyland and Hyland (2006) draw our attention to a fundamental mismatch between the teachers' goals and the students' expectations.Based on their research, they report that both parties ASIAN TEFL, Vol. 3 No.1, 2018, www.asian-tefl.com(teachers and students) have different orientations.The teachers in their research evaluated students' output only in terms of whether they were able to communicate the intended meaning while the students regarded teachers' feedback as a fundamental component of their language development, especially as far as accuracy was concerned.Hyland and Hyland (2006) posit that some teachers try to control the feedback process too rigidly and fail to take account students' needs and their goals in language development.Schulz (1996;2001), likewise, emphasize the presence of similar mismatches based on her studies conducted in the EFL context of Columbia and the ESL context of the United States.Ellis, (2009), in a similar vein, points out that learners typically prefer that the teacher provide the correction for them.Ellis also highlights that learners can only self-correct if they possess the necessary linguistic knowledge.
Self-correction is closely linked to "learner autonomy" as it is inspired by the following slogan: "Tell us, we forget; show us we remember; involve us, we learn."(Rief, 1990, cited in Sultana, 2009).Learner autonomy is usually used synonymously with varying other similar concepts such as 'learner independence', 'self-direction', and 'independent learning'.The SLA literature is replete with such slogans and concepts that are intended to captivate audiences, arguably with very little pedagogic value, for which (Widdowson, 1998: 705) offers the following comment: Although we as TESOL professionals say that we should not make changes for their own sake, we do make them for the sake of demonstrating that we are still dynamic, and for this purpose even the appearance of change will do.And this is why old ideas keep coming back with the veneer of novelty.But we do not want changes to be too disruptive either.It is preferable for our sense of security that they should be easily assimilated, and one way of managing this is to reduce ideas to simple terms that sound good: comprehensible input, natural learning, authentic language, whole language, real English.These become a kind of catchphrase currency whose value is taken for granted without further enquiry.And thus we become slogan prone.
Learner autonomy, in essence, is not only a methodological, but also an ideological orientation language, and to this end, it dictates that teacher intervention should be minimized, if not completely eliminated.It is ideological in the sense that it is a social process, which normally departs, or deviates from traditional education especially in terms of redistribution of power between the learner and the teacher as well as reassessing the status of learners in the process of the construction of knowledge (Masouleh & Jooneghani, 2012).Simply put, learner autonomy is a reformist orientation to learning which, in fact, attempts to fix learner behaviour.Esch (1996, p.37), however, emphasizes that "autonomy is not self-instruction learning without a teacher"; neither does it mean that any intervention or initiative on the part of a teacher is banned.In the SLA literature, autonomous learning is marketed in a way that it sometimes gives us the impression that it is synonymous with teacherless learning.The slogans used by the advocates of learner autonomy are so appealing that learner autonomy sounds realistic in terms of its goals it aims to achieve: "Helping foreign language learners to become autonomous is one of the fundamental and the universal duties of all of the foreign language educators have" (Yagcioglu, 2015, p. 428).
Most of the premises as well as pedagogical bases which underpin learner autonomy are fundamentally flawed.Swan (2005), in fact, argues that the principles and theoretical basis that underpin learner autonomy lack empirical support, and not many studies have been able "to demonstrate the lasting retention and availability for spontaneous use of features acquired by incidental on-line focus on form, and there are no reports of successful long-term classroom implementation of the approach" (p.379).Bruton (2002), in the same vein, posits that there is little evidence that tasks as core activities resolve the fundamental questions of accurate oral production.
As for the literature on self or peer-correction reviewed above, it seems that selfcorrection is not well appreciated by students since they need more support and guidance from their teachers.As Wright (2002, p. 115) puts it "a linguistically aware teacher not only understands how language works, but understands the student's struggle with language and is sensitive to errors and other interlanguage features".
As discussed above, some researchers argue that some fossilized patterns defy error correction.This may be true to some extent.If that is the case, then we teachers should abandon instruction altogether and allow our students to acquire the language naturally as some influential researchers argue (Krashen, 1982).My experience, however, points to the opposite direction.Having been a practitioner for more than almost three decades, I observe my students' linguistic behaviour carefully since I am aware that errors made by students give me an indication as to the incorrect patterns they produce may consequently fossilize.More specifically, errors made by students signal to teachers about the existing level or status of their interlanguage.Any teacher who does not reflect on students' output in general can be argued to put their students' language development at risk; one potential risk is fossilization, especially in the early stages of language development.As a practitioner, I do value and appreciate students' struggle with the challenging language learning enterprise, and to this end, I take into consideration every error seriously, reflect and act on it.I constantly provide feedback on their errors.The following incident took place in my class only last week.The student uttered the sentence: "He don't study hard."The moment he completed his sentence (without any feedback from the two teachers in the class), he said: "he doesn't study hard".From a pedagogical perspective, this is indeed a positive progress on the part of the student in that he had become aware that he and don't do not collocate well in terms of grammar.I can interpret this as a positive outcome of repeated and emphasized error correction.The preceding linguistic episode points to the fact that error correction is a must for teachers.The student's correcting his own mistake is encouraging since this is an indication that error correction works.Practitioners should patiently observe their students' errors, and especially watch out for frequently occurring errors.Obviously, error correction should happen irrespective of whether errors hinder communication or not.The ASIAN TEFL, Vol. 3 No.1, 2018, www.asian-tefl.comabove mistake (he don't study hard), would definitely not hinder communication as the meaning was communicated successfully.
To sum up the foregoing, this study attempted to investigate why Saudi EFL learners frequently produced the fossilized pattern "I am go".The hypothesis was that some learners produced "I am" to actually refer to "I".The findings obtained confirmed the hypothesis given that only one third of the participants (out of 155 participants) were aware that "I" was the pronoun that represented ‫انا‬ in Arabic.
This study aimed to fill a gap in SLA research considering that no research has been conducted on the fossilized "I am go" pattern.The research available only list and classify the errors made by Arabic-speaking EFL learners.Below is an account of few studies that limit their content to the categorization of the errors made by Arabic-speaking EFL learners.
Noor's 1996 article, in particular, is limited to listing the most common syntactic errors made by native Arabic-speaking learners of English.Noor (1996) provides a comprehensive list of seven syntactic categories which provides a detailed account of errors made by Arabic-speaking learners of English.He provides the following two examples that exemplifies errors relevant to the ones discussed in my study.
a-They aren't cause difficulties to us.b-I am not understanding the lesson.
Noor attributes the above fossilized patterns to aspect problem, and explains that students either omit the progressive (-ing) form from some verbs (as in Example a above) or incorrectly and unnecessarily add the progressive (-ing) morpheme when actually it is not needed as in Example b above.Since Noor focuses solely on the aspect problem, he limits his attention to the copula verb (the linking verb) and the verb that follows it, whereas my study focuses on the pronoun, which serves as the subject of the sentence, and the incorrect insertion of the copula verb "am" beside the pronoun.For the two patterns above, however, Noor does not provide a detailed explanation of the causes of the above fossilized patterns since his aim is limited to providing only a comprehensive list of errors that Arabic-speaking learners of English make.Noor's list of syntactic error categories contributes to the existing literature in that his list is indeed a very comprehensive and a well-structured one, and attempts to capture all the errors made by Arabic-speaking learners.
In her attempt to classify native Arabic-speaking learners' mistakes, Al-Ahmadi (2014) cites a similar incorrect pattern to the one in my study (I am go).Her example reads: "I am study English from the intermediate school"; she regards this error as the misuse of the Present Tense.Al-Ahmadi's work, similarly, is confined to the listing and categorization of errors made by Saudi learners of English and does not aim at an analysis of errors; neither does it address potential causes of the error as well solutions to overcome the incorrect pattern.Her judgment of the error as the misuse of the Present Tense, in my personal opinion, does not seem to be solid since the sentence upon which she forms her judgment (I am study English from the intermediate school) may rather contain an aspect problem (Present Perfect Progressive).The sentence can be corrected by saying: "I have been studying English since intermediate school."or "I have been studying English since I started intermediate school", implying that the error in the sentence is not actually an issue of the use of the incorrect present tense, but it is rather how the action of learning could be linked to a point of time in the past.Al-Ahmadi, however, regards this issue as learners' incorrect switching between the Present Simple and the Present Progressive.
In another article, Nuruzzaman, Shafiqul Islam, and Shuchi (2018) categorize errors committed by Arabic-speaking EFL learners (again in the Saudi context) and provide the following pattern as an example of a sentence structure problem.The sentence is: "I am go to university", which they correct as "I am going to university".The pattern in question is identical with the pattern in my particular study.The researchers interpret the pattern not as an overgeneralized pattern of the first singular person subject pronoun, but rather as a sentence structure problem; such a diagnosis however, in my opinion, is ill-defined.
If the sentence in question is assessed to have a structure problem only, such an assessment cannot claim to have captured the problem adequately since the assessment is too broad, and therefore not specific to the error in the sentence.
It is indeed the case that problems with tense, case, aspect, subject, or object can all render a sentence as a structurally problematic; that said, reducing the "I am go to university" pattern to a mere sentence structure problem does indeed not capture the error adequately.Such broad assessments of the problem, in fact, hinder an objective evaluation of the error in question.The best strategy to overcome such complex and challenging grammar issues should entail a process that involves a very specialized assessment of the error, which would tell us why exactly (factors causing the error) the error was made, and how it could be prevented.In other words, unless the error were assessed correctly, no solution could be devised to prevent the fossilized pattern from emerging.
Similar to the studies cited above, Nuruzzaman, Islam, and Shuchi also limit their analysis of errors to categorization of errors made by Saudi learners, and do not provide a detailed account of what factors cause the incorrect pattern and what action can be taken to avoid its fossilization.This implies that it should be an obligation for teachers and researchers to adequately define and specify the root of the error in order to develop viable strategies to battle fossilization.
As noted earlier in the Introduction section, most studies dealing with especially Turkish and Arabic-speaking learners focused on the fossilization of the definite article "the" (Snape & Kupisch, 2010;Abushihab, (2014); Al-Jamal, 2017).The fossilization of the verb "am" in particular has not undergone a serious empirical investigation in the Saudi context.This study, in this sense, can be regarded as being unique in that it does not limit itself to a sole categorization of errors, but rather tackles an error through an empirical and in-depth analysis.It also addresses a fossilized pattern (I am go) not in terms of the error in the verbs (am go), but rather in terms of the pattern containing a subject pronoun (I am).
Furthermore, the investigation conducted for this particular study was a pedagogical inquiry rather than a theoretical linguistics-oriented one given that the researcher investigated a phenomenon that he himself witnessed taking place due to his active role as an EFL practitioner.To the best of my knowledge, this is the first study that addresses the fossilized pattern "I am go" in such an analytic manner, which investigates the causes of the pattern and suggests pedagogic solutions to avoid the problem.
There are so many other fossilized patterns Arabic-speaking EFL learners produce (Noor, 1996;Abushihab, 2014;Al-Jamal, 2017;Noor, 1996;Al-Ahmadi 2014;Nuruzzaman, Islam, & Shuchi, 2018) that are observed in the Saudi context and elsewhere.Classification of errors might be a good start for error identification; however, researchers should go beyond classification, and analyze each fossilized pattern for the benefit of students.It is also critical that the classification, or categorization should be labeled correctly; unless errors are correctly diagnosed, as argued above, pedagogic interventions may not prove useful since the cause of the problem has not properly been identified.
This brings us to the conclusion that to facilitate pedagogy, we need to understand the precise causes of fossilized patterns.In addition, further research should be conducted with other fossilized patterns, especially by practitioners (EFL instructors).In particular, research on other pronouns (you, he, she, it, we, they) that may potentially fossilize with the insertion of a linking verb should be conducted to complement the findings obtained from this particular study.
The question remains as to whether we practitioners should limit our instructional happiness and satisfaction to our students' successful communication of meaning (oral fluency) with limited accuracy.Is it not high time we responded to the following question: "What is the real goal of second language teaching?"In order to answer this question, I think it is critical that we first respond to the question posed by Spolsky (2007, p. 10): "What does it mean to know a language?"In my view, to inquire about an individual's proficiency in a language, the most common and the most unlinguistic question, "Do you speak …….." (as in the question "Do you speak Chinese?") is disturbingly ambiguous as it reduces "language proficiency" to mere "oral proficiency", with little or no emphasis on proficiency in literacy skills, i.e. reading and writing.The SLA research, in my personal view, is severely disabled in responding to the latter question.It is indeed high time we integrated literacy skills into the so-called language "proficiency" concept, or devised a new question that would replace the ambiguous question "Do you speak ……….?" Figure 1