Reasons for Vocabulary Attrition : Revisiting the State of the Art

This paper reports on a one year, mixed-methods longitudinal case study investigating the neglected area of the perceived reasons why participants forget vocabulary knowledge. The participants were 43 fourth year male Saudi EFL majors at King Abdulaziz University KAU, Saudi Arabia. Quantitative and qualitative data including self-reported questionnaires and retrospective semi-structured interviews offered evidence to support the findings of this study. The reasons associated with lexical attrition centered on lack of practice, instructional and environmental context and nature of the word.


Statement of the Problem
A remarkable amount of research has been carried out in the field of language teaching and learning.However, one might pose the question of how much of the language skills acquired are maintained over longer periods of time when exposure to this knowledge is confined to the classroom.Closely related to the field of language teaching and acquisition research is research into language attrition, a phenomenon that has been a topic of investigation for the last three decades (Bardovi-Harlig & Stringer, 2010;Flores, 2010).Its beginnings can be dated to May 1980, when the University of Pennsylvania hosted the first conference ever on the theme.Most of the contributors to the conference called for systematic empirical research into language attrition which had at that time been overlooked.Further fundamental concerns have been raised by attrition researchers (e.g., Bardovi-Harlig & Stringer, 2010;Weltens & Cohen, 1989;Weltens & Grendel, 1993) concerning the variety of factors suggested as influences on the emergence of first, second and foreign language attrition in general and of vocabulary in particular.A number of variables have already been discussed in the attrition literature including personal or individual factors such as age (Berman & Olshtain, 1983;Cohen, 1989;Tomiyama, 2000) where severe attrition at the lexical level occurred in younger L2/FL learners; initial proficiency (Alharthi, 2012;Bahrick, 1984;Weltens, 1989) where the amount of attrition in vocabulary appeared to be independent of the peak attainment) although see Alharthi (2014a) for contrary results; and motivation and attitude (Gardner, Lalonde, Moorcroft, & Evers, 1987) whose findings failed to show significant correlations between motivation and/or attitude and tests of lexical production.A last group of variable is related to differences in individuals' claimed vocabulary learning strategies (VLS) where a reduced amount of English in rote learning led to more attrition in word knowledge by Saudi EFL graduate learners (Alharthi, 2014b).

Intrinsic Difficulty of Lexical Items and Attrition
At the same time, it has been suggested that there are a number of factors influence the learning of a lexical item and make the acquisition of vocabulary difficult.Potentially, these factors are classified as intra-lexical traits, i.e. internal features related to the word's form and meaning.That is, vocabulary may be difficult due to various factors, such as pronounce ability, grammatical class, similarity to known words, and concreteness or image ability of meaning (Nation, 1982).As long as there is a difficulty in learning a given word, it is reasonable to expect that a greater decline in vocabulary or forgetting of that word occurs.According to Laufer (1997) the internal factors that may impact the word difficulty and learning burden include pronunciation, form, morphology, syntactic patterns, semantic patterns and collocations.Additionally, Milton (2009) summarized intra-lexical factors that may have a facilitating or a difficulty-inducing effect on word learning including orthographic patterns, cognateness, concreteness and abstractness, word length and part of speech.The extent to which learning a lexical item is achieved by adult L2 learners might be accelerated further if the textbook would, in addition to the above effects of difficulty factors, take into account the salience of frequency in input (Milton, 2009;Schmitt, 2010).
The main purpose of the present study is to build on the findings obtained from a previous study (Alharthi, 2014a) in which Saudi EFL graduate teachers experienced drastic attrition in lexical knowledge, with the aim of contributing to the currently under-researched picture of the perceived reasons behind forgetting vocabulary knowledge.

Survey of the Relevant Literature
It should be noted that research has shown that vocabulary is relatively prone to attrition, more so than other language subsystems such as syntactic and grammatical categories (Al-Hazemi, 2000;Anderson, 1982;de Bot & Weltens, 1995).Nevertheless, to the best of the researcher's knowledge, little is known about the reasons for vocabulary attrition by Saudi EFL majors, and is thus in need of attention from applied linguistics researchers.Readers' attention, however, should be drawn here to the difference between forgetting and loss.Scholars such as Weltens and Grendel (1993) preferred to use the term forgetting to the term loss since the deterioration is not permanent but the vocabulary item simply inaccessible.One of the major theories that accounts for the forgetting of vocabulary learned in context is the "retrieval failure theory".Psychologists assume that forgotten language input is still available and not totally erased from memory (Loftus & Loftus, 1976, cited in Weltens & Grendel, 1993).It follows therefore that vocabulary is forgotten but not permanently lost.In the light of language attrition research, Cohen's (1989, p. 147) investigation of loss of productive vocabulary by bilingual children in Portuguese as a third language concluded that, "these words were not lost from memory but that the memory links were increasingly blocked by other interfering material, preventing the production of the desired words from one storytelling section to the next".
Empirical evidence from FL studies seems to suggest that Saudi learners experience the loss of certain words that they have learnt.A study by Al-Hazemi (2000) supports this conclusion; he found that Saudi officers in the Saudi armed forces forgot more than 50% of the items tested after leaving King Abdul Aziz military academy as a result of having no practice or enough exposure to the English language available to them.This finding seems consistent with Al-Akloby's (2001) study who found that even the summer recess was likely to have an impact on vocabulary attrition among Saudi high school students.One way of minimizing lexical attrition is likely to be via regular revision.Being loaded with many words to encounter and know requires revision in the light of insufficient lexical practice.It has been argued that when there is a lot to learn, there is a lot of forgetting (Cohen, 1989).Regular revision will help learners overcome a major obstacle in vocabulary learning and facilitates retention.That is, the difficulty of some vocabulary items such as being difficult to pronounce and spell, having several meanings, and being low frequency words can contribute to vocabulary forgetting by learners avoiding using them.This may force students to use general vocabulary as a compensatory strategy for communication but can lead to forgetting new words.Cohen (1989) reached a similar conclusion, i.e. that the choice of inappropriate lexical items can result in distorted communication, or linguistically incompetent L2 communication unlikely to be expected from EFL majors preparing to be EFL teachers.

Gaps in the Literature
Unfortunately none of the studies that investigated lexical attrition dealt with the neglected aspect of the reasons why EFL graduate teachers seemingly forget vocabulary.Hansen (2001, p. 61) stated that, "One might ask what impels this interest in language forgetting.On the one hand, language attrition research supplies empirical foundations needed for practice in language planning and language education."FL vocabulary attrition research may inform FL teachers of the long-term effects of their teaching, specifically the properties of word knowledge that are susceptible to attrition.Systematically investigating the reasons why EFL graduate teachers forget vocabulary may help practitioners and teachers to assess the immediate and long-term effects of controllable factors in instruction, such as syllabus, method, and length of study.Additional investigations of the causes of FL vocabulary attrition could help educators and researchers to provide EFL teachers with much more exposure to the FL through extensive reading and robust courses as part of their in-service teacher training.The study described here was conducted in an attempt to deepen the audience knowledge of the reasons for the deterioration of vocabulary attrition by Saudi EFL graduate teachers.I wish to reveal further individual factors which might affect attrition, for example, word length, word frequency, polysemy and specialism.With these considerations in mind, I set out to find answers to the following research question: What do EFL graduate learners themselves think are the reasons for their vocabulary attrition?

Overall Design of the Study
The present study had three parts, two of which are reported here in detail.Part A was purely quantitative in nature, with vocabulary tests to examine evidence of vocabulary attrition after the end of formal instruction.Parts B and C followed a mixed-methods approach in which self-reported questionnaires and semi-structured interviews were used to elicit and explore in more detail the participants' reasons for vocabulary attrition.Part A has been reported elsewhere (see Alharthi, 2014a).

Participants
The sample group was purposefully drawn from the quantitative investigation outlined in Alharthi's (2014a) study which initially consisted of 67 participants.Only participants who took part in baseline, second and third phases were included in the statistical analyses of this longitudinal study.At the final phase, i.e. two years later, there were altogether 43 participants since 24 individuals failed to complete the tasks during the second and third phases.

Instruments, Procedure and Data Analyses
Data to address the above research question were collected via a self-reported questionnaire and retrospective semi-structured interviews.That is, these two research instruments were used to provide information regarding the participants' reasons for their lexical attrition.The questionnaire was constructed by an awareness created according to the relevant research methods literature (Dörnyei & Taguchi, 2010).Specifically, the authors' guidelines for the structure and essence derived from the objective of the current study were deemed appropriate for the relevant context.The questionnaire started with a general question, "After the end of your English course, do you feel that you forgot some English vocabulary?" which had only yes/no response options.If the participants answer yes, then they proceed to the next step where they are asked to rate their opinions about each reason for vocabulary attrition in the form of a closed set of statements format in ascending order, ranging from 1 'strongly disagree' to 5 'strongly agree'.
Besides, an open-ended question was included at the end of special categories (e.g. if you agree, say what kinds of words; any other reasons, please say) in order to give the participants an opportunity to add whatever they felt, which might not be asked about in the questionnaire.In this way it was hoped to give participants a greater voice in the study.Participants were advised to complete open-ended questions in either Arabic or English.To address coverage, comprehensibility, feasibility and appropriateness, the questionnaire was piloted with a group of 10 EFL majors who did not participate in the main study.The numerical data were exported from the questionnaire into SPSS 20.0 (Statistical Package for Social Science).It should be born in mind that as the number of the final participants in Phase 3 (n = 43) was rather small, there was a possibility that it violated the assumption of parametric tests.For that reason it would be reasonable to employ non-parametric tests.To examine the significance difference between the participants' responses in the second and third phases, Wilcoxon Signed-Rank Test was used.I adopted the Bonferroni adjustment method (p = .017)to adjust the significance levels.The open-ended data were analyzed qualitatively according to recurrent themes in participants' responses.
The semi-structured interviews targeted the reasons the participants gave for the choices they made in answering the questionnaires.The interview followed the guidelines discussed in the relevant literature (Dörnyei, 2007).The participants were asked questions similar to the ones in the questionnaire in the interview guide to provide as much detail as possible and to enhance the target information to be obtained from the statistical analyses.Whenever needed, I had a great opportunity to probe and expand on topics of interest and relevance to the study.Since the interview items were entirely tailored to the questionnaire participants' responses to each statement, there were extracted and listed under such statement.

Quantitative Profile
The analyses and results using the Wilcoxon Sign-Ranks Test with the Bonferroni adjustment are shown in Table 1.Overall, the data in Table 1 show that all mean ratings (except "I think some kinds of words are easy to forget") on the items related to subjects' elicited reasons for lexical attrition decreased a little, curiously though significantly.As can be seen in Figure 1, most of the reported reasons were rated above the mid-point of the scale in Phase 3.This indicates that the participants in final sample are likely to agree about the possible reasons behind their vocabulary attrition.However, the pattern of responses in relation to the negative use of language in general and vocabulary in particular falls almost mid-way within the scale, suggesting at best that the subjects still practice the target language which is not surprising since they are already involved in EFL teaching.Interestingly, one EFL teacher reported that he sometimes bothers less about vocabulary when using the language, justifying this as follows: "I don't always focus on my vocabulary when I communicate with friends, e.g., if I am struggling to pick up a word and use it, and found an alternative or simply paraphrase what I am going to say, so most of these neglected lexical items are forgotten." The above extract supports the reason that vocabulary may not be given importance in general communication which leads to the speaker using more general vocabulary or paraphrasing as compensatory strategies to function effectively, but have an impact on the avoided items, leading them to be forgotten.
Comparing mean negative item ratings on revising/going over tests and words that had once been learnt at King Abdulaziz University (KAU), it can be seen that the study participants are more likely to see this technique as another reason for their lexical attrition.
Figure 1.Final sample participants' claimed reasons for vocabulary attrition over time In addition, the level of agreement tends to increase gradually in the participants' ratings of positive items regarding the remaining reasons of their vocabulary attrition between Phases 2 and 3.That is, they expressed their positive view of seeking out more different contents or vocabulary than the KAU input provided.This in fact creates two possible interpretations; one is that EFL graduate teachers are likely to take advantage of having access to various external sources for vocabulary that suit their career needs and life styles, which is expected since they are no longer studying the target language.The other possible interpretation is that seeking out more different contents or vocabulary might have a negative effect on attrition in the sense that the type of vocabulary input learnt in literature, linguistics and teaching method courses at KAU has not been given any attention since graduation; hence the words learned in these subjects are more susceptible to attrition.The following quotations from the interviews exemplify this aspect.
"Yes, I do need other vocabulary that is suitable for my teaching and everyday life.Those specific words I learnt at the college are forgotten, for instance, words related to psycholinguistics, language and culture courses." "Indeed I opt for words that cope with the nature of my work (teaching).Although I have been teaching at the university for three months I found words studied in literature, phonetics and applied linguistics courses are forgotten simply because you would not use these words in everyday conversation and writing." The participants still rated reliance on their L1 in EFL language teaching at the level of "agree" which is clearly positive in terms of the range of the scale in Phase 3. Again, this result is in agreement with the findings about the use of L1 in memory VLS, i.e., note taking and repetition (Alharthi, 2014b).This shows that EFL graduate teachers think that the inclusion of L1 in L2 classes is one of the possible factors leading to their lexical deterioration.It seems that EFL teachers feel that they ought to use their L1, which in turn prompt the question why do teachers decide to use the L1 even when they see it as a possible cause of their vocabulary attrition.The following quotations from the interviewees might provide answers to this question.
"One of the teaching situations I experienced before was when I started to speak in English, the whole class laughed at me.The fact that even if I use English I have to translate and explain the whole idea into Arabic so it is a wasting of time.Since then I have switched to Arabic due to the low level of their English." "Once I became a TA in the university, I thought I would practice the language all the time but I was shocked and found myself only using L1 in teaching 101 classes which might be the result of lack of motivation, commitment and background knowledge on the students' part.So my vocabulary vanished as time passed." What these quotations indicate is that Saudi EFL graduate teachers still adopt the grammar-translation approach as their normal means of classroom communication.The most agreed-upon reason for this was the low or limited English of their intermediate school students and the non-English majors studying at the university.As a consequence, EFL graduate teachers believe that this led to them forgetting their own vocabulary.
Curiously, the results showed participants increasingly acknowledging their using a limited range of words, and claiming that some words are easy to forget.The former makes sense because of the nature of their career, along with the teaching methods adopted and the content materials followed in classroom instruction, which might have an effect on the vocabulary repertoire being used in the restricted situations of classroom teaching.As a result, other learnt vocabulary is rarely revisited and thus neglected and is therefore in danger of being lost.Many interviewees shared this understanding as the following quotations show: "My vocabulary is more likely to be limited because of the primitive degree of lexical knowledge represented in the intermediate school materials".
"Most words I am using are guided by basic textbooks when teaching students at elementary and intermediate levels." Although a detailed discussion of the type and the level of lexical content in terms of thematic and frequency criteria in the school textbooks in Saudi Arabia is beyond the scope of the present research, there is tentative evidence, drawn from the above quotations, that course books used in public schools contribute nothing of any value to resist attrition in vocabulary.
The second reason given for attrition, i.e., that some words are easy to forget, was given the highest frequency rating by the study participants.A follow-up question was included to find out the types of words the subjects thought are easy to forget.Drawing on both questionnaire and interview data, the subjects' responses given in Phases 2 and 3 are grouped Table 2 by their lexical characteristics.The data shows that five main factors are responsible for increasing the learning burden and leading to lexical deterioration.Predominantly participants found that word length, word frequency and subject-specific words are likely to be implicated in attrition.Parts of speech were also implicated in lexical attrition, specifically adjectives were more likely than verbs to be lost.Further, words with a single form but several meanings were more likely to be forgotten according to two study participants.
The same linguistic features were identified by some individuals in the retrospective protocols, as shown in the following quotations: "There are a number of aspects of vocabulary learning that affect the process of attrition, like words of more than one syllable, e.g., globalization.Also, words which belong to the existing courses at KAU, e.g., psycholinguistics are mostly forgotten." "Difficult academic lexical items, e.g., the ones I learnt in psycholinguistics and literature courses at KAU like 'cognition' and 'plot'.I have difficulties in word spelling therefore; long words are always prone to forgetting." "I find it difficult to communicate in particular situations, e.g., in hospitals and airports where they require specific vocabulary to function effectively.These are more easily forgotten.Parts of speech is another issue in my case, e.g., nouns are resistant to attrition while verbs and adjectives are more vulnerable to attrition.In contrast, I found long words more memorable, e.g., 'congratulations' since it is widely used." All the questionnaire responses and quotations from interview above confirm the views of EFL learners noted in the literature of difficult L2 words and their learning burden, e.g., Nation (1982), Laufer (1997), Milton (2009) and Schmitt (2010).Laufer (1997) suggests that factors such as a word's grammatical class (adjective/verb) and length (long words) can make learning more difficult and therefore make these words less likely to be retained in memory.As for the role of part of speech (POS), the present results may give strong support to the study's quantitative findings that nouns are more resistant to attrition since they are easier to learn (Alharthi, 2014a).However, it might have been hard to predict the conclusion with respect to the length effect.As reported by one participant, quoted above, long words are not necessarily easily forgotten, e.g., congratulations, since they are commonly used in spoken language.
This again agrees with Laufer's (1997) alternative argument that word length has no clear effect on the ease or difficulty of learning a word but it is rather the learner's frequent exposure to it that is of importance.This highlights the weakness of intuition about the word length effect, since word length is not always a reliable sign of attrition.It is interesting to note that the results of the current study are in line with the conclusions by Milton (2009) and Schmitt (2010), which is that specialized lexical items, used in restricted situations, are more likely to be forgotten.Lastly, different meanings of a word, other than the basic one, seem less likely to be memorized, a view supported by two cases as shown in Table 2.
These results go a long way to help demonstrate that certain word categories, such as specialized or polysemous ones, are predictive factors in the attrition process.However, since they represent only a small number of cases in the present study, they should not be taken as definitive.

Open-Ended Questions and Interview Data Related to Reasons for Forgetting English Vocabulary
An open-ended item was included in the questionnaire, prompting the participants to speak more freely about the factors which contribute to their vocabulary attrition.All participants attributed their lexical deteriorations to the poor FL environment in Saudi Arabia, believing that they do not practice sufficiently inside the classroom as EFL teachers and outside the classroom as L2 learners.This supports the results discussed in the previous section about the reliance on Arabic by most Saudi teachers in the EFL class.The following quotation from an interviewee illustrates the limitation of L2 input in Saudi context.
"The learning environment in Saudi Arabia itself constitutes the problem which causes some attrition of English although I could practice it though only to limited extent."Some responses given in the interviews revealed that the degree to which English language is practiced varies in different environments and regions in the country, depending on people's culture, which negatively affected the participating EFL teachers' language use, as shown in the following quotation.
"The practice of English differs from one part to another due to some public school students coming from families with poor literacy and education, and therefore you would not expect them to use English, and not even have the desire to learn English, so their language and vocabulary is not resistant to attrition." This quotation is drawn from an EFL teacher who teaches English to intermediate school students in Albaha, a small town in the southern region of Saudi Arabia.He said that vocabulary is vulnerable to forgetting as English does not receive importance as a language to learn in EFL classes, and that this is due to some settings, especially small towns in most regions of Saudi Arabia where most students come from lower class families with poorly developed literacy and a lack of appropriate education.Thus, the extent to which the language is used in teaching is rather limited as students are clearly less motivated to learn and use the language.This observation is to some extent in line with Al-Akloby (2001) who found that students studying at secondary level received little help from their parents to learn the language, probably because the latter did not have sufficient education themselves.
Another participant reported much the same, explaining the limited facilities available, specifically limited internet access and authentic materials in libraries provided by EFL institutions which lead to attrition of his lexical knowledge.
"One of the possible factors contributing to my language and vocabulary attrition are the workplace conditions under which I am working in Taif University, which lacks facilities like computers, printers, materials available in library and most importantly internet access." The above quotation revealed that the facilities provided by educational institution impact the achievement of this EFL teacher and ultimately his language/vocabulary attrition.The Ministry of Higher Education in Saudi Arabia has provided audio recordings, computer laboratories, up-to-date technology and internet access as instructional facilities in most universities and colleges around the country.Nevertheless, some branches and campuses under the supervision of the main institutions lack these facilities.
The above results confirm that language environment, i.e., the setting where the target language was acquired, is one of the influential factors which have been previously demonstrated to be relevant in the context of language attrition (e.g., Weltens, 1987;Weltens & Grendel, 1993).In this respect Weltens (1987, p. 22) states that, "the attrition of a particular FL may differ both quantitatively and qualitatively in situations where the post-course exposure is of a different kind.In other words, certain environmental characteristics may partly prevent language attrition, others may speed it up".Along the same lines, the findings of the current study are consistent with Al-Hazemi (2000) who found that military officers forgot more than 50% of the words they had learned at King Abdulaziz military academy which was mainly due to the lack of an appropriate FL environment.

Conclusion
The article has presented a detailed discussion of the quantitative and qualitative results of an investigation of EFL graduate teachers' possible reasons for their vocabulary attrition.The results show that there are many different kinds of reasons perceived by the study participants to contribute to their lexical attrition.The reasons given to some extent overlap, hence they are grouped under three main categories: • Factors related to lack of language practice or to less focus on lexical aspects when using the language.
• Factors related to instructional and environmental contexts: (i) type and level of materials used by the study participants in teaching classes; (ii) teaching methodology used to deliver the target language, for example, by using L1; (iii) lack of motivation by public school students to learn the language, possibly due to their parents' poor education, perhaps even illiteracy; and (iv) lack of adequate facilities provided by some institutions.
• Factors related to the linguistic 'nature' of a word, e.g., length, frequency, specialized use, polysemy and part of speech.

Pedagogical Implications
The current study pinpoints a major cause of language attrition that the participants do not get enough exposure to vocabulary, which surely is not good enough, particularly for EFL majors.The EFL environment of the participants has a remarkable influence on their lexical retention.Lack of daily use of language vocabulary such as in or out of class interactions with colleagues, and with native or native-like speakers, contributes to lexical forgetting, and a decrease in vocabulary.The fact that the participants in the final sample of EFL teachers reported that they keep little contact with English suggests that in many EFL settings access to any kind of spoken English might be severely limited.Teachers and materials writers should guide L2 learners to move on to a level of additional lexical knowledge through extensive exposure in reading materials, because of time investment.In other words, classroom vocabulary input mediated by teachers could be supplemented by having them read extensively, which helps to commit their vocabulary to memory.Schmitt (2008) believes that the key to EFL vocabulary enhancement and retention is to integrate extensive reading with explicit enhancement activities.It is of paramount importance that the language teaching policy at KAU maximizes the opportunities for using language in more meaningful communication practice, especially for graduates who are preparing to be EFL teachers.
Most importantly, EFL graduate teachers can later take steps to foster their language skills in general and their vocabulary knowledge in particular that can be maintained after formal instruction ends.This can be through undertaking FL reading, such as books related to their areas of interest, authentic texts (such as magazines and newspapers), watching TV, listening to radio and browsing the Internet, so that either existing or new lexical items will be met again and again and so will be retained in memory.Evidently, as Nation (2001, p. 155) argues, "the use of reading and other input sources may be the only practical options for out of class language development for some learners", which is especially true for EFL environments.
It might be helpful to add another strand to their language development as a part of in-service training courses.I suggest that such courses could serve not only with regard to instruction in teaching techniques, but also empower EFL teachers with linguistic skills, particularly in vocabulary, speaking and writing, and so encouraging them to participate in group discussions.In this way, teachers are furnished with the chance to maintain the vocabulary they have already learnt.In fact, the duration of an in-service training course is likely to have a significant impact on long-term vocabulary retention if it is delivered on a weekly basis.

Table 1 .
Wilcoxon analyses of reasons for vocabulary attrition in Phase 2 and Phase 3

Table 2 .
Linguistic factors making it easy to forget lexical items, given by EFL teachers in Phases 2 and 3