This is a two part work. The latter part, Appendix, refers to work carried out in 1997-1999 in relation to the Critical Age Hypothesis. The first part relates to a subsequent critique of the research methods employed in the earlier work.

A1. Introduction.

A2. Initial research.

A3. Overview

B1. Critique of Abstract

B2. Critique of Literature Review

B3. Questionnaire

C. Critique of - Validity and Reliability of the Research

i. Descriptive validity

ii) Interpretive validity

iii) Theoretical validity

iv) Internal validity

v) External validity - reliability

vi) Reflexivity

D 1. Other General Issues if the Research Were to be Repeated

E. Conclusion

Appendix 1. Abstract.
Appendix 2. Literature Review.
Appendix 3. Survey overview.
Appendix 4. Survey.
Appendix 5. Results.
Appendix 5. Survey conclusion.
Citations.

A1. Introduction:

A couple of years ago a rare opportunity in EFL/ESL education annals occurred in Korea. The English curriculum, which hitherto had begun teaching English in Middle schools to students at the age of 14 changed to a situation where students in Elementary schools would begin to learn English. So for one year we had two groups, namely 8 year olds and 14 year olds beginning to learn English for the first time. This presented a unique opportunity to survey students whose age coincided with the critical period hypotheses debate. The schools are all part of the Korean government education system. The question at the heart of the research was whether young Korean learners found learning and the speaking of English easier than older students and adults, and secondly, if teaching English pronunciation to 8 year olds had educational development merit.

A2. Initial research:
The material the subject of this critique is annexed at page 12, (Abstract) page 14 (Literary review), and pages 23-46, (Survey questions and analysis). This is in the original form as prepared in 2000 and 2001.

A3. Overview:
Influencing this research is the issue of Confucianism and how it impacts on the Korean students. J.K. Lee (2002:45-61), arguing for the Confucionist survival notes that contemporary Korean education policy is influenced by an intricate combination of factors, dominated by Confucianism. The policy translates to practice and W. Lee (1996) states "…Asian students are not only diligent, but they also have high achievement motivation" but these factors may have less to do with Confucianism than other factors, such as, according to Ko, (2002) being "…educational zealots." W. Lee's view (1996), finds support from J. Lee (2002:58) who talks in terms of the "…educational enthusiasm of Korean people…" Ellinger and Beckham, (1997), note, "South Koreans view education as they view the rest of life: a process of winning and losing. They have no concept of a game played well for its own sake. The family emphasis on educational achievement is so strong that it has been dubbed "education mania.""
Whichever view one subscribes to, Confucianism with its attendant rules of filial piety, or educational enthusiasm or educational zealots, or a combination of both, it is clear that the issues, to some degree or other, impact the subject of this research.
It became arguably clear that qualitative research, "...the systematic collection, organization, and interpretation of textual material derived from talk or observation" Malterud (2001:483), was applicable. Against this was the consideration of ethno- methodology research, described by Cohen, Manion and Morrison, (2001:24) as "…the mechanisms by which participants achieve and sustain interaction in a social encounter - the assumptions they make, the conventions they utilize, and the practices they adopt." The key words in this definition being, 'the conventions they utilize, and the practices they adopt.'
As culture, Confucianism, and an arguably different attitude on education exist in Korea, then a survey employing the caveat of Saville-Troike (1989:110), viz, "the need for an extensive background study of the community is critical, and a variety of field methods must be employed..." formed the basis of the underpinning research methodology. It is clear that the research needed close attention to issues of validity, as opposed to reliability, for the applicability of this to other scenarios is restricted given the uniqueness and unrepeatability of such a survey. That is not to deny reliability per se, but that the same set of circumstances will be rare to find again, but micro situations could be assessed for similar results. Thus only general conclusions going to reliability can be drawn from this survey research, for it is Korean and Confucian specific. However, this research, being in a field of intense debate, will be attacked over issues of validity, hence the attempt to limit the attack and to argue that any attack must be seen in light of ethnographically considered features the subject of this fieldwork.

B1. Critique of Abstract.
Whilst the abstract discussed the issues surrounding the critical period hypothesis debate, and the research at hand, one issue overlooked that may have been addressed was the differences, if any, between an EFL and ESL country. Kroll (2001) addresses the differences that result in different style of teachings, as does Matsuda (1998), however, it seems a deeper consideration of 'culture' and 'EFL' needed elaboration.

B2. Critique of Literature Review.
The question at hand is one that generates much debate, research criticism, and voluminous journal and book writings. As well, the area is developing in a different direction as proponents of MRI scans produce new data from their research. The literature review states all positions from a definite existence to there being no such thing as a critical period for language learning, to the mid point that that says both yes and no. However, once the topic was decided upon, namely do Korean students perceptions align with the proponents of the debate, then it became necessary to "…become familiar with the information available on the topic..." (Christensen, 2001:109). However, it seems that a simplified review would have sufficed, for the debate is well known.

B3. Questionnaire.
Given that the research question centered on perceptions of English learning, and to "…provide an answer to the research question…" (Christensen, 1997:64) then question 6 and question 16, which being general in nature, should have been removed from the survey.
The questionnaire used closed questions with a proscribed range of answers from which the respondents could choose. This was decided upon as an open question requiring writing would have produced high level of stress in students whose L2 written level was low. This could have been altered for the adult teacher respondents, however, again it was feared the response would be muted, as the teachers may have perceived the answers as testing their English skills.

As Cohen, Manion and Morison, (201:249) note, "…ambiguity in words is intractable: at best it can be minimized rather than eliminated." This presented a critical problem with the survey, namely, the skills of the Korean English teachers in explaining the questions, and secondly, whether the questionnaire should have been in Korean or English. At the time, it was decided to use English only, as it immediately sensitized the students to the questions at hand. Had the survey been done again, I would have opted to write the question in English, followed by a Korean translation, with the responses being in English only but orally explained by a bi-lingual teacher.

C. Critique of - Validity and Reliability of the Research
i. Descriptive validity
Descriptive validity refers to accuracy in the reporting of descriptive information. (Burke-Johnson, 1997:283). It is suggested this definition extends to the explaining of certain things, in this case, the questionnaire. 5 Korean English teachers and myself attended at 6 schools and collected the data. Whilst it is clear field investigators need training, this survey involved the explanation via bi-lingual speakers of the questions the subject of the questionnaire. It is submitted that each field worker was a form of corroboration for the others, except in the case of my field work, for my bi-lingual explanations did not reach the same level as fluent bi-lingual speakers, nevertheless, corroboration comes from a variety of field methods, (Saville-Troike, 1989:11).

ii) Interpretive validity
Interpretive validity is defined as "…the degree to which the research participants' viewpoints, thoughts, feelings, intentions, and experiences are accurately understood by the qualitative researcher." (Burke-Johnson, 1997:284). Only in the case of adult participants (Korean teachers) was feedback sought. Their views were obtained pursuant to ethnographical criteria (a long time within the community and using pre-testing samplings and feedback) (Saville-Troike, 1989), and such that critical cultural norms would not be breached. Pre testing member checking ensured the questions posed were non-offensive and did not breach the standards of a Confucionist society.

iii) Theoretical validity
One of the precepts of the survey involved 4 years in-situ in Korea learning and observing Confucian customs. The questions thus considered issues locale' specific and were peer reviewed prior to the survey commencement. However, peer review during data collection may have highlighted issues needing corroboration or clarification.
iv) Internal validity
"Internal validity refers to the degree to which a researcher is justified in concluding that an observed relationship is causal" (Cook and Campbell, 1979). Whilst Burke-Johnson (1997:288) notes the importance of listing questions that show rival explanations or rival hypothesis, having identified a causal relationship, Fox (2000:22) notes the critical importance of asking the right questions to obtain the best data possible. The relationship has long been shown as the younger the L2 learner, the greater the L2 output proficiency. Rival causal explanations form the basis of the debate. In this survey, a variety of external influences were noted and built into the testing. If the survey were to be repeated then clearly researchers would be needed to consider a variety of issues that would take time and money to research, arguably centering upon the effect of Confucianism as an internal or external factor influencing the causal connection.

v) External validity - reliability
Could the results of this research be applied to other groups? As results confirmed the literature review that pronunciation programs are best initiated with the young L2 learner, then the results, considering the issues raised in this 'perceptions study' could definitely be applied, however, as mentioned above, issues of EFL/ESL/Confucianism need further consideration, which was not done in this survey, however would have provided greater reliability.

vi) Reflexivity
Biases that go to the validity of the research are "…the characteristics of the interviewers, the characteristic of the respondent, and the substantive content of the questions", Cohen, Manion and Morrison (2001:121). Not only do the words of the questions asked orally become important but also so do the words of written questionnaires, (discussed above).
Clearly the research required greater transparency about the author's background (substantiating ethnographic considerations) and greater transparency about the Korean English teachers who assisted in the research. However, as observation did not play a part of this research, but was substantially questionnaire based, then issues pertaining to culture in society needed further elicitation, for it was a contention of this research that hitherto research had done too little when considering the beliefs and feelings of the research subjects, as opposed to the research that had simply looked at differences between a young learner and an old learner. Thus biases about culture are definitely a possibility and need careful elicitation.

D 1. Other General Issues if the Research Were to be Repeated.
(a) One of the dilemmas associated with the research related to ethical issues. There were three levels of dilemma. First involved carrying out the research in a government school. School principles unanimously agreed to the research. The second level of dilemma was involving the senior Korean English teacher in each school to participate. All teachers agreed to participate, however the sheer size of the survey meant that that teacher lost the majority of one lesson.
The final level dilemma was the students themselves. Students were presented with the questionnaire as a fait accompli, and were simply asked to complete the questions. No parental permission was sought. However, anonymity was guaranteed; "...information provided by the students should in no way reveal their identity", (Cohen, Manion and Morrison, 2001:61). Whilst the students undoubtedly found the break from the monotonous lesson routine to be enjoyable, and arguably ethnographic considerations were carefully covered, the sheer amount of questions probably angered some teachers immediate post survey who assisted in data collection. One thing not done, that should have been done, was feed back to the teachers who assisted in the research as to the results, plus a small thank you note.
As opposed to the students who had no information supplied before the questionnaire was filled in, Korean teacher respondents to the survey were free to undertake or not undertake the survey, and were fully informed pre-survey as to the reasons and uses the data would be put to. It was noted that some teachers did not proceed with the survey.
(b) The amount of questions, whilst raising some interesting issues, were clearly too many. Despite peer review pre-survey, which suggested about 8 questions should be asked, a total of 16 were asked. However, according to Cohen, Manion and Morrison (2001:259), what should have been done was to have broken the question sequencing down, to give the appearance of simplicity and manageability.
(c) Data analysis.
Data was collected from 6 schools across the metropolitan area. It was assimilated into one batch and results were aggregated. If the research were to be repeated, I would change this process so as to (i) analyze individual school results (ii) compare the schools to each other for consistency or inconsistency (iii) then aggregate the data to compare an aggregate against the individual school data.
However, it should be noted that this was in a manner similar to the post-survey survey of 50 students, some 3 months later, with students chosen for one school (not one of the original) whereby the results were significantly similar to the original survey.
(d) School selection.
Whereas the schools chosen for the Yates (1999) research survey were split between metropolitan and rural zones, the schools the subject of my research were all metropolitan schools based in a city of nearly five million people. The locations of the schools did represent different socioeconomic zones and were schools that were classified as A, B, or C class schools. However, the addition of a rural school would have been a useful comparison.
(e) Statistical comment
After each question percentage data comes an analysis. The views need further in depth analysis, and more analytical consideration in light of the ethnographic concerns expressed above.
(f) Statistical analysis
Whilst data was presented in simple %s, as no appropriate software was available, a further consideration could present the data somewhat differently, as bar graphs to highlight similarities and or differences.

E. Conclusion:
The size and expectations of the survey were quite both large yet realistic. The results were not meant to add to the debate that considers traditional views versus the neuro-linguistic developments, rather than present data that may assist the debate by considering ethnographic, psychological and place specific cultural criteria. Indeed, the results, taken at a humanistic level, did show a different perspective on the critical period hypothesis from an individual's view. The survey also statistically showed that the widely held view that pronunciation is one area that should be taught to young L2 learners is arguably correct and with sound basis.
The initial questions asked what could have been differently done had the research been repeated. Some issues were presented above, each having its own weight of importance. Given that the uniqueness of the situation that occurred will never happen in this country again, it waits to be seen if something similar occurs in another ESL country. Finally, to reinforce the caveat of Saville-Troike, (1989:110), a greater variety of field methods would have been a tool going to greater survey validity.

Appendix 1.
Abstract:
Korean classrooms, until the beginning of the 7th Curriculum, 2002, had traditionally taught English via rote learning contrary to Confucionist principles that dominate Korean education. Little or no attempt was made to place the element of English speaking pronunciation in any perspective. The literature and previous research on pronunciation programs is conflicting, and depending upon which approach or methodology the curriculum follows, either ignores it per se, or leaves it to implicit assumptions about how to teach an acceptable level of pronunciation to Korean school students.
One thousand six hundred Korean students and one hundred and two Korean teachers of English were surveyed on their perceptions and beliefs about the English language and learning good pronunciation in Korean schools, this being an EFL country.
Secondly, a rare global opportunity existed for testing the critical age hypothesis in that Korean learners' of English started at two different ages. The lower age fell within the limits of the lower end of the critical age hypothesis (8) and the second group of new English language learners fell within the upper limits (14) of age for that hypothesis.

This research suggests that, in the case of the Korean government school classroom learner, that students' perceptions about learning a good English speaking pronunciation fall well within the critical age hypothesis debate on the side of the proponents affirming its existence, yet does suggests that the time frame for learning good English pronunciation, as opposed to supporting directly the debate, suggests that a window of opportunity exists within which to compile successful pronunciations programs for the Korea learner.
The research also suggests that as Korea is an ESL country, a new Approach or Methodology specifically for an Asian country practicing Confucionist principles is needed, and that traditional methods and approaches are not applicable for specific EFL countries with specific cultural differences. English pronunciation programs for Korean students need specific tailoring, taking into account the window of opportunity for learning good English pronunciation, and teachers who can teach English pronunciation.

Appendix 2.
Literature Review:
Critical period hypothesis.
The critical period hypotheses propounded by Lenneberg (1967) suggests that primary language acquisition occurs during a critical period which ends at about the age of puberty and must occur before cerebral lateralization is complete, and the follow up implication being that second language acquisition will be relatively fast, successful and qualitatively similar to first language only if it occurs before the age of puberty.

The classic argument for this proposition is that there is a critical point for second language acquisition and that this point is around puberty. "… beyond which people seem relatively incapable of acquiring a native like accent of the second language." (H Douglas Brown, 1994, 93) This theory has led some people to assume that once past the age of 12 or 13 you are incapable of successful second language learning. Brown (1994, 56) goes on to say, "Some adults have been known to acquire an authentic accent in a second language after the age of puberty, but individuals are few and far between."

However he fails to explain what is an "authentic accent" in this multi-cultural English-speaking world. Bickerton (1967) and Lennenberg (1981, 53) argue for the critical age hypothesis. O'Grady and Dobrovolsky (1996, 464) say in relation to first language acquisition based on research studies, "It is now widely believed that the ability to acquire a first language in an effortless and ultimately successful way begins to decline from age six and is severely compromised by the onset of puberty." They go on to conclude that the answer to whether there is a critical age period is answered by both 'yes' and 'no.' (O'Grady & Dobrovolsky, 1996, 491) Their conclusion also indicates that there is nothing biological that prevents adults from acquiring proficiency in a second language, whereas H. Douglas Brown (1994) indicates that research on the critical age theory, shows brain lateralization affects second language acquisition, thus opposing the view of O'Grady and Dobrovolsky (1996). Scovel (1988) in his research suggests that "…plasticity of the brain prior to puberty enables children to acquire not only their first language but also a second language and that possibly it is the very accomplishment of lateralization that makes it difficult for people to ever again easily acquire fluent control of a second language…"

Walsh and Diller (1981:18) found that different aspects of language are learnt at different ages. "Lower order processes such as pronunciation are dependant on early maturing and less macro neural circuits which make foreign accents difficult to overcome after childhood…" Thus this research gives support for the critical age thesis in so far as certain areas of language (pronunciation) come inside its parameters. However they left open the conclusion that obtaining a fluent foreign accent is not possible after puberty. Klein (1995:24) seems to temper the argument by saying, "…cases of adults acquiring a second language without any accent are very rare. This alone does not provide sufficient evidence that second language learners are in principle unable (for biological reasons for example) to attain a native pronunciation." This supports the Walsh and Diller line of argument. Neufield (1978) suggests that it is possible for adults to obtain native like levels of language proficiency. Ellis (1996) makes a broad proposition by saying most L2 learners fail to achieve native speaker level ability. Later he appears to contradict himself by saying only "…child learners are capable of acquiring a native accent…" but qualifies this to informal settings.

Singleton (1989) concludes that massive amounts of exposure are necessary for a child to achieve native like proficiency, yet Ellis (1996) concludes even massive amounts of exposure still results in children failing to achieve a good level of proficiency. The reason, he concludes, is their desire to maintain an active use of their mother tongue. However this seems to overlook the issues of inhibitors.
According to Seliger (1978) there may be both 'critical periods' and 'sensitive periods' for language acquisition. Generally the view favored is that pronunciation and native like proficiency precedes the period for grammar acquisition. But according to Thompson (1991) starting early is not a guarantee of achieving native like proficiency, no matter how favorable the teaching system is. McLaughlan (1992) argues that the younger the learner the more skilled he becomes in that language is a myth and disputes the findings of Krashen, Long & Scarcella, (1992), yet concedes that, "Pronunciation is one aspect of language learning where the younger is better hypothesis may have validity;" Asher & Garcia (1969) concluded that the younger the learner the more native like accent that student develops. McLaughlan (1992) suggests (following the line of Scovel 1982) this is because "…pronunciation involves motor patterns that have been fossilized in the first language and are difficult to alter after a certain age because of the nature of the neurophysiological mechanisms involved."


To further his theory over the uncertainty of the whole debate, he suggests that it may be that educators simply do not know how to teach phonology in a second language, impliedly suggesting that teachers of a second language need further and updated specific skills. This 'radical' proposition may in fact be closer to the 'ugly' truth that would see the critical period debate thrown out wholus bolus on the basis that the true answer lays in the professionalism of the educators and nothing more! Should this theory gain hold, the EFL profession would see a radical change in its criteria for teaching methodologies and teacher credentials.

What McLaughlan (1992) also suggests is that because children have to learn shorter and simpler structures in their early acquisition days, there is an 'illusion' that children learn a second language more easily and quickly than adults, yet testings have shown the exact opposite to be the case.
Studies of age differences in the acquisition of second language pronunciation have produced conflicting results. Snow and Hoefnagel-Hohle (1977) found better pronunciation in older students, whereas Fatham (1975) found a negative effect occurred with age. Seliger however makes reference to scattered cases of adult learners acquiring accent less pronunciation in their second language, and conversely some children who had maintained foreign accents. Professor Stapp of the University of Tsukuba, Japan provides further valuable insight with research of 28 monolingual Japanese students repeating lists containing the letters /r/ and /l./

Her research noted, "The ability of young children to achieve native like proficiency in a foreign language in a rather short time is a reflection of a type of neural plasticity, and it appears to be related to the distinct characteristics of the young brain."
Her research sets out to establish whether young learners are better at 'mimicry' than the adult learner based on neural plasticity properties. Early studies by Tahta, Wood and Lowenthal (1981) suggested that superior performance of children's mimicry was due to neural plasticity that "…promotes native like L2 pronunciation in early acquisition." Long (1990: 266) supported this line of argument, viz; "The sharp drop in imitation abilities …after age 6 can be accounted for by positing that maturation constraints begin to set in as early as 6 for supra-segmental phonology in some learners and soon after that for segmental phonology…Phonological attainment is strongly conditioned by learner age."

But as Stapp notes; "An assumed relationship between neural plasticity and the ability to mimic raises two important problems. First, not all young children are good mimics: but statistically, young children do outperform older children and adults in L2 production over the long run. Second, some adolescents and adults are good mimics. However, since the type of neural plasticity that accounts for superior L2 pronunciation in early childhood actually declines with age, it is unlikely that such plasticity is responsible for the outstanding mimicry skill exhibited by relatively few mature individuals."
Medical research has shown that first language, including simultaneous bilingualism, is densely represented in the language areas of the left-brain hemisphere, a second language acquired later is typically more distributed in the brain. (Ojeman and Whitaker 1978) Further investigations by Kim, Relkin, Lee and Hirsch (1997) confirm that second languages acquired in very early childhood are spatially close to first language in the frontal lobe, while languages acquired later are well separated in this region. This research is still in its infancy as to definitive conclusions.

Flege (1987) rejects the notion of critical periods, and argues that neural plasticity affecting second language acquisition is not subject to time constraints, but full L2 acquisition is possible at any time. Jacobs (1988) furthers this line of argument with more persuasive arguments.
Neufeld (1978:1979) agrees. His research demonstrated the permanence of neural plasticity, thus eliminating the critical age. He says that a 'sensitive period' is a false notion. Klein (1986:10) takes the issue into the social sphere; "The biological explanation for difficulty in L2 acquisition after puberty can be replaced or supplemented, by arguments of a social nature. …ideal second language acquisition is biologically feasible even after the age of puberty." This is furthered by Lowenthal and Bull (1984) who suggest that language proficiency is boosted significantly by an encouraging environment.
Stapp concludes her research by noting that mimicry skills appear to be a distinct talent distributed across the age population, thus eliminating the connection to neural plasticity. Her corollary states that; "If mimicry skill is not related to neural plasticity in children, it is probably not possible to invoke neural plasticity as a basis for superior performance by adolescents and or adults.

However, "the evidence clearly indicates that the typical child has the advantage over the typical older learner in the ultimate acquisition of L2 pronunciation, with or without good mimicry skills at the outset." Nevertheless, the mid point position suggesting a critical age period as opposed to sound contradictory evidence, is of great importance to elementary school teachers in Korea who are responsible for the 'critical age children.'
Of interest and concern and encompassing the critical age theory are two letters of the English alphabet that the Koreans have the perceived greatest difficulty with, namely the letter "L" and the letter "R". Douglas Brown (1994:57) notes that an infant with developing speech muscles advances from simple cries to articulated vowels and consonants by the age of 5, however, "…complex sounds …r and l…sometimes are not achieved until after the age of 5, though virtually complete phonemic control is present in most 5 year old children."

This has serious implications for Korean students trying to master the pronunciation of the letters "l" and "r". This will be returned to later. According to Yamada and Tohkura, (1992) it is especially difficult for Japanese children to pronounce the letter 'R' after the age of 14. However, learners vary in multiple other ways apart from age, and it is these differences, associated with Terrell's Natural Approach (Krashen), which will subsequently be looked at and linked together in returning to our central research questions.

However, the Korean teacher at this level (elementary) has studied and received his/her qualification mostly from a University of Education where broad stroke training is applied, namely all topics are learnt with not one specifically, hence Korean elementary teachers do not have specific language training skills in the English language, for it is but one of 12 or more topics they must teach. The Korean Middle school system is for 3 years. English is taught in all grades. The difference here is that the teachers have majored at a University in English and their knowledge and spoken level of English is supposed to be better than the Elementary school teachers. The teachers also only teach one subject, namely English.

High school is for 3 years, with teachers possessing the same educational ability as the Middle school teachers, however, as spoken English is not an major part of the High school English system, some Korean High school teachers can't speak English and if they do it is of a sub-standard level on the whole, such that if 'understandable English' is the criterion, they fail the test. However as the High school exam system is not geared towards spoken English but only grammar and rote memory, the High school teacher needs rarely speak a word of English in his class. This is undergoing change however as University entrance exams will now place emphasis on spoken English, thus one can foresee a demand for competent English language teachers in High schools. Hence, in Korea, if we accept the critical age theory, then an urgent reassessment of just who teaches English and what are the qualifications, is vitally important, for logically, by age 12-14, the student who now has a qualified teacher, is in the situation where it is too late to benefit from that teacher's skills. (also see Conclusion chapter)

Recent research by scientists at the University of Alabama into this vexed question have broadly determined that in all skills, the maxim the 'younger is better' is applicable. They further conclude that the more words a child hears before the age of 2, the better vocabulary he or she will be able to develop later in life. (Matire. J. 1999, 95)

In conclusion with other factors traversed in this book, it is suggested the term, "windows of opportunity" is more apt than 'critical period.' If one accepts the word 'critical' then one impliedly accepts that teaching a second language learner after age of 14 is a lost cause. This presents psychological factors on both the teacher and student side, which could affect the training. It is submitted that if we consider this time frame (2 - 14) merely as an excellent window of opportunity to teach our pronunciation program and teach according to principles applicable to that age range, we will succeed more often than not in our teaching program. However, merely because we have missed that window of opportunity, doesn't mean categorically we can't teach pronunciation successfully. What it means is that the teacher will have other factors to consider and employ in his teaching program, and that success in teaching what we can term as acceptable pronunciation is still highly probable, and that we should also make the student aware that our goal is such and is attainable. This is the sign of a good language teacher.

Appendix 3.
Survey overview:
Much debate exists over the likelihood- existence of Critical Age period for SLA. (noted in depth herein.) Much of the debate is hypothetical or based on small survey samplings. In 1998 the Korean Education system offered a totally unique existence to make some comparisons. Prior to 1999 Korean students only began to study English in Middle Schools (age 14) for the first time. Those students who had received any English education prior to their first year in Middle school were omitted from the survey.

Thus at the beginning of the survey we had the rare situation of children beginning English language lessons at Middle School for the first time, and children at Elementary school (aged 9) also learning English for the first time. The starting point was to obtain as large a sampling as was possible. To this end, Korean elementary students from 3 schools were surveyed. Children in third grade elementary were sampled. A total of 602 children from elementary school grades were sampled. Also students from 3 Middle schools were sampled; one boys school in what is considered a low socio/economic district where one parent families are not unusual, and a second girls only Middle school in an average socio/economic suburb. The Middle schools are single sex schools whereas Elementary are co-educational. 800 students were sampled at these schools.

A further small sampling was obtained from a private school, where students are of adult years and pay to attend to receive their education. Thus the sampling came from government and private schools. Finally a survey of 102 Korean English teachers was taken.
The age range was from third grade elementary 9 years old and upward. Of course it must be pointed out that Korean age is two years older than western age for the same age, i.e., when a Korean child is born he is already one, and come January first he turns two, hence a child born in December 1999 becomes 2 in January 2000.

The survey was presented in the written English language, with guidance from a bi-lingual Korean English teacher to assist the students with unknown words. The reasons for presenting in the target language were; a) responders immediately associated with the language they were being asked about; b) responders were then required to respond with their acquired second language, thus, though not identifiable, arguably increasing the coefficient of reliability and affective factors towards genuine responses.

The following factors were considered present at the time of survey that could influence any result.
a) Elementary schools.
a) time of day survey taken. Elementary students become progressively more tired after lunch than in the first period. It is anticipated a tired mind negates motivation and could affect the response.
b) Weather. The survey taken was conducted in the autumn, when daily temperatures were reaching the mid to upper 20s Celsius, with high humidity present. Some days the weather provided oppressive conditions in the classroom. This may negatively affect enthusiasm of either doing the survey or motivation to the target language.
c) Interpretation. The survey, in English language, was translated by a Korean English teacher to the class. Obviously a key factor is the interpretation of the words and the understanding of the question by the students. A further factor was the closeness of the translation to the question framed.
d) Class atmosphere. Some classrooms showed visible signs of 'color and friendship' towards the English language. Others remained bleak. Thus class motivation and mood towards English must be present.
e) Socio economic regions. The survey was taken over 6 schools covering what was considered to be a fair range of living standards.
f) Affective factors. At the time of survey students had been at school for a second term period of some four months, with still almost 2 months of study remaining. Student's motivation, it can be presumed with no need for analysis, would be less than say in the final week before a long summer or winter break.
g) Classrooms in which the survey was carried out were; either in new modern luxurious schools, or in old and 'dirty' schools.
h) Elementary schools are well equipped in each room with the latest multi media facilities, computer, dvp, etc, thus given an overt look of high quality training.

Middle schools.
Some factors that could affect the response are.
a) Middle school students are generally stressed and tired at all times owing to large amounts of study. English may be seen as a 'burden rather than something 'living' and enjoyable.
b) Single sex schools. Unlike elementary schools, middle schools are single sex. The survey covered one girl's school from a middle class district to a boy's school in a lower class factory zone.
c) Second language learning in middle schools is more book and drill oriented than in elementary schools, where teachers are generally more inventive and less restricted in terms of curriculum schedules, thus students are more likely to be 'tired' and physically exhausted than elementary students.
d) Classrooms were generally devoid of any affective materials and gave a cold inhospitable feel and look.
e) Classrooms were also devoid of multi-media equipment, adding to the negative atmosphere.

Korean Teachers. A sampling of (102) was taken.
Some factors that could affect any objective result noted were;
a) the survey was taken on day two of their teacher training courses. The course began at 3.00 p.m. after the teachers had completed a long working day. Tiredness was a factor here.
b) Korean teachers have suffered a substantial pay cut following the IMF crisis and it is expected this will in some 'negative' way affect their answers.
c) Teacher's level of comprehension. Despite the fact they were teachers, it does not follow their English was at a level that could understand and respond to the questions correctly in some circumstances.
d) Teachers surveyed at teacher training classes were compulsorily conscripted to attend.
Thus, in an attempt to spread the 'negative and positive factors around, surveys were carried out during the first and last lesson times, in classrooms where the temperature was either ideal or slightly uncomfortable.
Further, in the case of the student survey, only classes surveyed were those where the teacher interpreting the survey was a senior English teacher at that school, so as to reduce misinterpretations as much as possible. However, a caveat exists; (see Gieve (1991) in relation to a survey conducted in Japan that noted that differences in responses were discovered depending upon whether the questions were presented by a Japanese teacher or native English-speaking teacher.

As Ellis (1996) says, doubts must exist as to whether such survey results can be valid and reliable. However, little short of responders answering on a computer-generated survey, the element of non-native or native presenter cannot be overcome.

In this Korean survey, a range of socio economic zones was surveyed, ranging from affluent to poor. A further point common to all groups is the countries overall view of the English language. At the time of survey, and unlike during the IMF crisis when anti western literature was appearing daily in the media accusing the 'west' of manipulating the crisis, at the time of survey, only one negative press release accusing western nations of manipulating the current economic situation appeared, although some sentiment was expressed against the US involvement in Serbia, not to mention an anti American attitude running through society in relation to U.S. troops on Korean soil.

How does one build in a factor to take into account all these issues? Maybe all that can be said is that the test had general face validity. (Hughes. A. 1991, 27) Whilst suggested analysis's for a 'second testing' were not considered feasible, for in such a survey, certain factors present in one survey may well be removed in the next, or complemented by corollary issues not hitherto accounted for, nevertheless, small random second samplings provided similar results fitting within the + or - 3% variable. It can only be said that by surveying a wide range of students and adults in a variety of Affective conditions, can one somewhat equate the results to a medium point.

In the following questions, both question 14 and question 15 leave themselves open to the Social Desirability Effect phenomena. An example in this survey can be seen in question 14, where respondents may answer "yes" despite the fact that their local Korean teacher speaks excellent English without a foreign accent, and produces just as good a pronunciation model as the native speaker. One way considered to reduce this possible survey defect was by using forced choice question such as, "Whom do you think it is better to learn pronunciation from?
(i) a native English speaker
(ii) one of your Korean English teachers who speaks very good English."
However, even this question is not likely to produce perfect data because the phrase "…to learn pronunciation…" is too imprecise.
Some respondents will address the fact that a native speaker will, all other things being equal, tend to provide a more accurate model of correct pronunciation, and therefore will choose (i). However, others may be more mindful of the fact that the Korean teacher is able to explain the niceties of English pronunciation to them in their own language, something most native speakers are unlikely to be able to do, and therefore will choose (ii).
In other words, different respondents may interpret the question in different ways, and the data gained from their response will therefore not represent the true views of the group surveyed. Having noted this, it was felt, nevertheless, to leave question 14 and 15 in their raw form.
Question 6 also uses the word 'pronounce' and does not make any distinction between spontaneous learning, that is free from any language tutoring or second language schooling and the second type, namely systematic learning. Nor does it impose any time limit living in the target country. However, the base question is free of issues that could confuse the responder, and is in response to recent media, television and print media that has strongly suggested that Koreans need not travel to another country to learn fluent English. (Korean Herald)
Question 17 was added following suggestions from many Korean teaching staff.

Appendix 4.
SURVEY:
16 questions were asked. The first four are purely statistical:
1. Age.
2. Sex.
3. Number of children in the class.
4. Years study.
5. Have you lived in an English speaking country?
6. Do you think living in an English speaking country would help you pronounce English;
much better
a little better
I can learn in my own country
7. Which letters of the alphabet do you find difficult to pronounce?
8. Do you like speaking English?
9. Do you speak English at school, home, both?
10. How do you learn English pronunciation?
- teacher
- watching t.v.
- listening to music
- other
11. Why are you learning English?
-school rule
-parents want me
-I want to
-my friends are
-for future or work
-other
-it's fun
12 Which 3 foreign languages (after Korean) are the most important to you learn?
- English
- Japanese
- Chinese
- German
- Spanish
- French
- other
14. Can you learn good English pronunciation from a native English speaker?
15. Can you learn good English pronunciation from your home country (Korean English) teacher?
16. Omit from report.
17. Do you think it is
a) very important
b) important
c) useful
d) no good
in learning body language movements as part of a pronunciation and communicating course?

RESULTS.
The survey was broken down into three components.

Component A. Elementary school students in their 2nd and 3rd year of English language study. The age was 11 and 12 years old.
Component B. Middle school students in their 2nd and 3rd year of middle school. The age is 14-15-16 years old.
Component C. Korean teachers and adult learners of English. The age range is from the mid 20s to the upper 50s. The average age was 29.
By breaking the survey down to 3 specific life zones, with an interval of four years minimum between Elementary and Middle school students, and a minimum of ten years interval from middle school to the youngest teacher, it was hoped that changes would be observed in the time spans.

**There is also one considerable advantage and benefit through this method. The Korean education system is in flux. We have the rare situation of Elementary school students beginning to learn English (Korean age 10-11-12) western age (8-9-10).
Similarly we have Middle School students who have only begun learning English at Middle school, and have had two to three years study. Their Korean ages are (14-15-16) Western ages 12-13-14, thus we have rare insight into the Critical Age theory, from the early days in the critical age, to the later days. The survey reflects a unique window of opportunity to examine students at two ends of the spectrum both doing the same thing.**

Appendix 5.
ELEMENTARY RESULTS. Survey group = 880 students. E
MIDDLE SCHOOL RESULTS. Survey group = 820 students. M
TEACHERS. Survey group = 102 T

Question 5. Any student who responded affirmatively was not counted in the survey.

Question 6
Do you think living in an English speaking country would help you pronounce English;
E. MS. T.
Much better 63% 62% 51%
A little better 24% 26% 26%
I can learn good pronunciation
In Korea. 11% 11% 22%
Comment.
There is a small falling off (much better) in belief as the age increases, yet still a substantial proportion believe in the virtues of going to study overseas. Business operators can take heart in this and should consider their advertising campaigns! It also shows the Korean authorities have failed in their attempts to convince the students that they can receive a good level of pronunciation training in their own country. In January 1997 there were 133,249 Koreans studying abroad in 69 countries. The IMF era has seen this 'exodus' (Korean Herald October 29, 1998) decrease substantially, although according to Matire. J. (1999, 29) that will only be temporary, especially with countries such as the U.S.A. easing regulations and allowing Korean students to hold part time jobs whilst studying. Note the editorial, Chosun Ilbo Newspaper, 11th June 2000, strongly advocating Korean students save their money and study English in Korea as opposed to Canada. This may have changed post September 11.

Question 7.
Which alphabet letters in English do you find difficult to pronounce?
E MS T
l (light -dark/approx) 11% 12% 50%
r 10% 34% 92%
f 9% 27% 70%
p 18% 10% 27%
z 34% 60% 89%
T / D 34% 76% 91%
Comment.
N.b. Phonetic symbols were provided on the blackboard by the Korean English teacher for all sounds.
A significant difference is noted in Elementary 'r' and Middle school 'r' with the difficulty factor increasing. Given that there is no continuity of education, we cannot say the student's perceived an increase in language difficulty. However, we can say that at the end of the 'critical age' students with no formal training show markedly higher rates of perceived difficulty as they are entering second language acquisition at the end of the critical age time frame. This consequently must have serious implications for all teachers at elementary level, who must initiate pronunciation programs into their training.
A second significant perceived difficulty arises with the letters 'r,f,z, and /T/ D/ sounds. This has serious implications for teaching pronunciation, such as minimal pairs, sound glides, etc, at Elementary level.

Also it must be noted that with the exception of the letter 'p' all letters showed (for elementary and middle schools) an increase in perceived difficulty. This in itself is worthy of further investigation, namely given the critical age period has elapsed, why has the letter 'p' gone in the opposite direction to the others.

What seems to be stunning is the percentage rate of perceived difficulty with adult Korean teachers. The significant figure is the rate between /l/ and /r/. It is consistent in elementary schools. But both Middle school students and adults show a marked difference in perceived difficulty with the letters.
This indicates strong support for the proponents of the critical age hypothesis. It is known these adult teachers did not receive formal pronunciation or communication training at school or teachers college. It is also known their ability has been largely self-attained well after leaving school, indeed after leaving Teachers Training college. If anything, it shows the urgent need to formulate a clear and decisive training course including the fundamentals of speech and pronunciation given that they themselves are responsible for Elementary school English training. Conversely however, it may just show an adult's desire to master perfection in the target language.

Question 8.
Do you like speaking English?
E MS T
Yes 68% 33% 34%
No 7% 47% 43%
Not sure 22% 19% 22%
Comment.
It seems as the Elementary student proceeds to Middle school, the inhibitive factors increase, resulting in an increase in those who don't like studying English. This result should be considered next to point 7 above, with the increase in perceived pronunciation difficulty. Of significance is the high percentage of those who remain undecided about their beliefs on learning English.
Also over the age range it can be seen that those who like speaking English decrease with years, those who don't enjoy speaking English similarly increases. The reasons need further elaboration to determine if inhibitive factors have 'selective' periods and are more common to one group than another. It was noted ((not listed in the above analysis)) that a greater percentage (62%) of boys showed a positive preference to speaking English than girls at the Middle school level. This point may need further in-depth analysis.

Question 10.
How do you learn pronunciation?
(Only one category is included)
E MS
From schoolteacher. 67% 75%
This result, although open to arguments of being too narrow in scope, merely indicates that the teacher plays a perceived significant role in the teaching of pronunciation.

Question 11.
Why are you learning English?
E MS
i. school rule - -
ii. my parents want me to 10% 10%
iii. I want to 37% 21%
iv. my friends are 37% 10%
v. for my future 5% 50%
vi. other 89% 10%
vii. it's interesting and fun 54% 10%
Comment.
Students were allowed multiple answers hence the mathematical anomaly.
School rule was answered almost 100%.
Of note is the relatively small percentage where parental influence applies, despite earlier discussions on this point herein.
Sadly the self-desire to learn falls heavily from Elementary school to Middle school reflecting factors of concern, namely what has caused the fall off in interest? Similarly peer pressure falls away markedly in the Middle schools, yet the concern for the future soars from Elementary to Middle school. This does not sit comfortably with the small percentage of parents who want their middle school child to study English. It would be assumed that parents would discuss the future employability of their child with their child. The alternative is that schoolteachers have installed this belief (fear) in the students whilst they are at Middle school. And despite H. Douglas Brown's argument (1996) that peer pressure from friends to learn English exists, this seems resoundly refuted by these results. The final point of concern shows the massive decline from Elementary school where children enjoy English to Middle schools where they do not. This must reflect on Middle school curriculum and or teaching and practices and needs serious attention.

Question 14.
Can you learn good English pronunciation and good spoken English from a native English teacher?
YES E = 79% MS = 56% T = 86%
NO E = 20% MS = 43% T = 13%
Comment.
It seems surprising that a large percentage of all categories think it possible to learn good pronunciation from a native English teacher, especially as it is noted elsewhere teachers have few qualifications, (something private schools could exploit) yet only about 50-60% believed (in point 6) that they could significantly improve their pronunciation by living in an English speaking country. The discrepancy, it is submitted, is possibly (probably) due to the fact that Koreans believe they have a 'platform' of proficiency in pronunciation above which they cannot go. That platform, it is submitted, can be achieved in Korea from a qualified native English teacher with a clear pronunciation program.Although the figures for those who do not believe a native English speaker can teach pronunciation well are 13% -43%, it is submitted this figure is consistent with those who indicated in (point 8) above, that they do not like speaking English. It is submitted that that category has a negative factor built in about liking and learning English. This may be consistent with Porter & Garvin (1989) who warn that pronunciation programs may do more harm than good and should be further researched.

Question 15.
Can you learn good English pronunciation and good spoken English from a Korean English teacher?
YES. E = 55% MS = 37% T = 53%
NO. E = 9% MS = 15% T = 17%
Not Sure E = 35% MS = 47% T = 29%
The figures for Elementary students and teachers on the 'yes' vote is consistent, whilst the figures for the Middle schools is also consistent with the fact that they have only had two or three years English study, and fall on the outer limit of the critical age period, thus making learning that much harder and their perceptions more negative and although those teaching them have majors in English from a University, the spoken English of Korean Middle school teachers often leaves much to be desired, thus the added perception by the students that English is all the more difficult because their teacher either does not speak English (as happens in an English class, or the level is sub-standard, thus making learning that much harder.)
Similarly, many Middle school students have had tuition at private schools thus exposing them to native English teachers, and allowing a comparison to be made. Interestingly, less than one quarter of all categories believe that Korean teachers can't teach good English pronunciation, although the undecided factor is high. Clearly, however, the Korean education system has a lot of work to do in the field of training Korean English teachers if it is to meet the student's demands and expectations.

Question 16.
Do you think it is
a) very important
b) important
c) useful
d) no good
in learning body language?
a) E = 25% MS = 23% T = 19%
b) E = 44% MS = 23% T = 25%
c) E = 25% MS = 34% T = 53%
d) E = 0% MS = 6% T = 2%
Comment.
As argued above, non-verbal gestures (body language) falls within the broadened definition of what is pronunciation. The figures in all categories provide a good positive spread indicating that 'non verbal communication' should be taught. Of note is the tiny percentage that does not believe that body language is important at all. If anything, this stresses the need for curriculum writers to consider inserting sections and exercises relating to body language. This undoubtedly, if well done, would provide a fun and interesting way for students to learn about another aspect of this area of English, and indeed give them something tangible to compare to their Korean (native) language.

Survey Error analysis.
A sample survey of 50 elementary students, 50 middle school students and 20 teachers was conducted 3 months later outside of the survey time frame. The results fell within a + or - 3% difference in all categories, suggesting the figures obtained from the initial applicants were fairly indicative.

Appendix 5.
Survey conclusion.
Applicants from a broad spectrum were surveyed and surveyed under a variety of conditions and from a variety of socio-economic areas.
The results tend to give weight to the hypothesis of the critical age period, not as an absolute irrefutable time frame which beyond all is impossible, but as a time frame in which it is highly ideal to professionally and correctly teach students English and pronunciation. This time frame I prefer to label, "The window of Opportunity."Also, the results give weight to the theory that affective conditions are very significant. Further, the results show that there is far more work to be done, firstly in the teaching of pronunciation to students, secondly, in determining why such a large fall in those 'who like English' occurs after elementary school, thirdly, that English as a language seems to fail when it comes to public relations, in so far as a great percentage of applicants showed results suggesting a negativity towards the language. The survey does show, however, that more questions and more research needs to be done in relation to affective conditions.

Index:
Amtmann, J. & Evans, R. 2001, 'Interview Protocols', Corrections Compendium, vol. 6. pp1-11.
Asher. J. & Garcia. R, 1969, 'The Optimal Age to Learn a Foreign Language', Modern Language Journal, vol. 53, pp.334-341.
Bickerton, L. 1967, Roots of Language, Ann Arbour, Karoma Publishers. NY.
Brown, H. Douglas, 1994, Principles of Language Learning and Teaching, 3rd edn, Prentice Hall, NY.
Burke-Johnson, R. 1997, 'Examining the Validity Structure of Qualitative Research', Education, vol. 118. pp.282-286.
Caporaso, J.A. 1995, 'Research Design, Falsification and the Qualitative - Quantitative Divide', American Political Science Review, vol. 89. pp.457-461.
Carr, W. & Kemmis, W. 1985, Becoming Critical. Education, Knowledge and Action Research, The Falmer Press, London.
Cheon, H. 2001, The viability of computer mediated communication in the Korean secondary EFL classroom. Ma Thesis, Pukyong National University, Korea.
Christensen, L.B. 1997, Experimental Methodology, 7th edn, Allyn and Bacon, Boston.
Cohen, L., Manion, L. & Morrison, K. 2001, Research Methods in Education, 5th edn, Routledge,