Etymology and Vocabulary Development for the L2 College Student

November 1999 — Volume 4, Number 2

Etymology and Vocabulary Development for the L2 College Student

Tom S. Bellomo
Daytona Beach Community College, Florida


The result of an in-class experiment demonstrated that etymology as a word attack strategy for L2 students does not offer an unfair advantage to students whose L1 is Latin-based (the Romance languages). A significant portion of the English language makes use of Latin roots and affixes; therefore, students whose original language is derived from Latin will initially have an advantage in word recognition due to cognates. A pre-test did in fact demonstrate this disparity between the two populations, though near equivalent scores on a post-test demonstrated that both Latin-based and non Latin-based languages equally benefited from direct instruction which utilized etymology as a word attack strategy.


Foreign students are being denied access to credit courses due to their inability to pass mandatory tests designed for incoming freshmen at various community colleges, specifically, the reading and writing components of the CPT (Computerized Placement Test). Since proper placement has been associated with lowered attrition (Webb, Gay, Rhoden & Tripp, 1988) numerous colleges over the past decade have implemented mandatory placement exams to determine which students are in need of remedial course work. In an effort to enable foreign students to overcome this hurdle and expedite college credit enrollment, one particular college had instituted an ESOL college prep writing course targeted exclusively for the L2 student, though such a class had not been developed for reading. When L2 students began doing poorly in the heterogeneous (L1 & L2) reading class, a proposal was made to research and develop a separate ESOL college prep reading course. Research pointed to schema theory (Chen & Graves, 1995; Steffensen, Joag-Dev & Anderson, 1990) and vocabulary acquisition (Nation, 1990) as major components particular to the L2 learners’ needs. [-1-]

Of interest to course development was the question “How can vocabulary acquisition be ESOL specific?” which in turn pointed to etymology. With nearly 50% of the English language being derived from Latin-based words and many others derived from Greek (Smith, 1995), etymology was chosen as a word-attack strategy which focused primarily on Latin-based [LB] vocabulary. It was at this point that course development was challenged when colleagues disparaged the use of etymology, claiming that the use of LB word parts would pose a disadvantage to those whose native language was non Latin-based [NLB]. Their rationale suggests that French, Italian, Spanish and other Romance languages are derived from Latin; therefore, the LB student’s vocabulary would proximate the target roots, affixes and vocabulary.

An experiment was devised in hope of offering empirical evidence demonstrating the following hypothesis: The use of etymology is an efficacious word attack strategy regardless of the student’s original language background. Though Latin-based students will have an initial advantage due to their vocabulary schema (cognates) as evidenced by a pre-test, knowledge gained through direct instruction will ultimately yield relatively equal results, as evidenced by a post-test. This would imply that etymology as a word attack skill does not unduly favor one population over another.

For the purpose of this report, etymology does not refer to the study of word origins (historical), but refers to comprehending vocabulary through the knowledge of roots and affixes.

Review Of Literature

Recently, there has been an increased awareness of the role of vocabulary acquisition with regard to the foreign learner (Zimmerman, 1997). However, sources are scarce once one factors in the use of etymology and its comparative effect among different language groups. Contemporary studies in reading and L2 vocabulary acquisition often cite the work of Michael West (1953). His work is highly relevant to reading and second language acquisition; providing a background of his work is beneficial since his word list is still in use today.

In October of 1934, a conference sponsored by the Carnegie Corporation was held in New York City to discuss the possible use of English as a world language. It was agreed that in order to expedite English language acquisition, the foreign student would benefit most by becoming familiar with those words most commonly encountered in text. In 1939 Michael West was commissioned to revise the 1936 Interim Report on Vocabulary Selection. The onset of World War II delayed this task, but by 1953 he had assembled “A General Service List of English Words”, comprising the 2,000 most frequently used words found in English text. This task was accomplished by compiling a running text of five million words taken from a broad selection of reading material; words were then tabulated on the basis of frequency. [-2-]

Nation and Newton (1997) stated that this list “account[s] for at least 85% of the words on any page of any book no matter what the subject matterþ. Focusing learners’ attention on the high-frequency words of the language gives a very good return for the learning effort” (p.238). These words should be learned to the point of automaticity (Coady & Huckin, 1997). Jim Coady, graduate chair of the Department of Linguistics at Ohio University, suggests that the General Service List should not be a subject of direct teaching, but is best learned incidentally through students’ repeated exposure to text. He recommends that students be immersed in low-level text of their choice and interest in order to acquire these sight words (personal communication, March 25, 1999). Breland (1994) and other contemporary researchers cited West’s work as a breakthrough and still highly relevant, though it was assembled some 50-plus years ago. Where researchers showed concern, however, was in the inclusion of some items which are no longer used as frequently today, and conversely, other words which are popular today and were not being used, or used as often, at the time of the compilation. The example of “racism” was cited, which is not on West’s list, yet through heightened awareness during the 1960’s “racism” has since become a high frequency word. In a personal review of the text it should also be pointed out that this list was published in England; as a result, an occasional word (e.g., “smite”) occurs which would not be relevant to the language student studying in the United States. A less problematic point to consider would be the variant spelling of British words, e.g. “color” (American) versus “colour” (British).

Subsequent to West’s work was the formation of the University List (Nation, 1990), a list of 800 additional words, which was conceived under the same premise but incorporated words one would find in collegiate texts. This “academic language” accounts for an additional 8% of frequently encountered vocabulary; hence this composite list of 2,800 words will yield the reader an understanding of about 93% of the vocabulary found in almost any text. Another 2,000 words of the technical nature will yield a further 3% of the most commonly read vocabulary, but it is recommended that these words, which are course-specific, be reserved for study within the various content areas. The remaining 4% of all English words make up an enormous 120,000+ vocabulary items. These are low-frequency items, which coupled with their extensive number, make it impractical to learn by direct teaching. Instead, various word attack skills should be employed to derive meaning from these unknown words. To summarize, foreign students should first concentrate on high frequency vocabulary. Once these words are adequately understood, students are now ready to learn and apply word attack strategies in order to unlock the meaning of unknown, less frequently occurring words.

Barbara McGavin (1990) created an English curriculum organized around LB derivatives. Students were assigned a target list of eighty common Latin roots along with rules governing word analysis, but this study was not specific to the L2 learner; in addition, her sample was a high school class, not adults, and comprised eighteen students. Her report, however, demonstrated a positive correlation between the use of etymology and vocabulary growth. [-3-]


A quasi-experimental design was produced making use of pre and post-test scores. The convenience sample comprised an ESOL college prep reading class. The initial study utilized 11 participants from a total of 17 students. Five scores were not recorded since these students enrolled after the first week, and therefore did not take the pre-test at the same time as their peers. An additional student’s withdrawal at midterm fulfilled the mortality threat to validity. Of these eleven students, three were of a Latin-based language origin (two Hispanics and one French); non Latin-based languages comprised Russian, Estonian, Turkish, Vietnamese, Chinese, and Japanese.

The pre-test was administered at the onset of the fall, 1998 semester. It was multiple-choice (four answers per question) comprised of three parts: 10 questions asked students to identify a root’s synonym, 5 questions pertained to prefixes, and 15 vocabulary words. The test was collected, and students were not permitted to see their results so as not to give them an opportunity to review or study their errors, or to enhance recall of the test items. A post-test was administered twelve weeks later. The test was replicated during the spring term of 1999; nine additional students, none of whom were [LB], composed this second group. A second replication was given during the fall term of 1999; this provided an additional sample of seven LB students and seven NLB students.

Results from all three classes provided a total of ten Latin-based students: Spanish, French, and Portuguese; and twenty-four non Latin-based students from nine different countries (see Appendix).


The numbers were manageably low; therefore, conclusions were drawn by comparing the mean and standard deviation between the two groups.

Table 1. Overall Test Results

(n = 24)
  Mean SD Mean SD
Pre-test 60.7 8.94 43.54 13.17
Post-test 89.3 6.98 85.13 13.61

One NLB student, who eventually failed the class, scored unusually low on the post-test; that one score diminished the group average and skewed the SD. Eliminating that student’s post-test score would alter the NLB Mean and SD to 87.22 and 9.16, respectively. [-4-]


The small size of this convenience sample prohibits one from drawing strong inferences or making a firm prediction. Results, however, were in keeping with the hypothesis and should warrant further replication until a sufficient sample is derived.

What preliminary insights can be gleaned from comparing these test scores? First, as expected, the LB group performed better on the pre-test; this author suggests that cognates were most likely responsible for this apparent advantage. However, once instruction began both populations were simultaneously being introduced to new course material, and they were now individually responsible for the vocabulary items being presented to them. They were both learning the meaning behind numerous roots (112) and affixes (83), which in turn held them accountable for over three hundred and sixty new words; not included in that number are the numerous derivations formed by the suffixes which were introduced in the course.

Notice at the top of the NLB list, a Russian student scored 20% on the pre-test; additionally, a Vietnamese student scored the same. Those same students worked conscientiously and diligently throughout the semester and subsequently posted scores of 100% and 93%, respectively. In fact, exactly half of the NLB sample scored above the LB mean. Second, post-test averages between the two groups are very similar, and if the one student’s substandard score of 37% were factored out, the average of both groups would be nearly identical: 89.3 [LB] and 87.22 [NLB]. These results show that any advantage initially evidenced at the beginning of a semester by one population eventually flattens out to similar results by the end of the term. This suggests that both classes of students have learned equally well from implicit vocabulary instruction utilizing word parts.

To summarize near-equivalent post-test scores between the LB sample and the NLB sample suggests that LB students do not have an advantage over NLB students when learning vocabulary by etymology.


Empirical research and self-report surveys have borne out that vocabulary acquisition is one of the more important needs of the foreign adult student learning to read English. This experiment has taken a first step in demonstrating that etymology may be a viable word attack strategy useful for a college level, heterogeneous ESOL reading class, irrespective of a student’s L1.

Replication of this experiment is encouraged in order to provide an increased sample size. [-5-]


Breland, M. H., and others [sic] (1994). The college board vocabulary study. College board report no. 94-4. ERIC Document reproduction services no. ED382928.

Chen, H. -S., Graves, M. (1995). Effects of previewing and providing background knowledge. TESOL Quarterly, 29, 663-682.

Coady, J. (1997). L2 Vocabulary acquisition a synthesis of the research. In M. H. Long & J. C. Richards (Series Eds.) & J. Coady & T. Huckin (Vol. Eds.), Second language vocabulary acquisition. The Cambridge applied linguistics series. (pp. 273-290). New York: Cambridge University Press.

McGavin, B. (1990). Augmenting language analysis skills through Graeco/Latin derivatives studies. (microfiche: no.: ED324686).

Nation, I. S. P. (1990). Teaching & learning vocabulary. Boston: Heinle & Heinle.

Nation, P., Newton, J. (1997). Teaching vocabulary. In M. H. Long & J. C. Richards (Series Eds.) & J. Coady & T. Huckin (Vol. Eds.), Second language vocabulary acquisition. The Cambridge applied linguistics series. (pp. 238-254). New York: Cambridge University Press.

Smith, E. L. (1995). Contemporary vocabulary. (4th ed.). New York: St. Martin’s Press.

Steffensen, M. S., Joag-Dev, C., & Anderson, R. C. (1990). Reading Research Quarterly, 15, 10-29.

Webb, N.A., Gay, M., Rhoden, D. A., & Tripp, J. (1988). Computer adaptive assessment at Central Piedmont Community College. Accuplacer User’s Notebook, College Board, New Jersey.

West, M. (1953). A general service list of English words. (new impression: 1980). London: Longman Group.

Zimmerman, C. B. (1997). Do reading and interactive vocabulary instruction make a difference? An empirical study. TESOL Quarterly, 31, 121-138.

About the Author

Tom Bellomo has been in professional education for 12 years. He received his BA in Linguistics/TESOL from State University of New York, Stony Brook and his MA in Language Arts from the University of Central Florida. He is currently a faculty member at Seminole Community College, Sanford, Florida. He taught in Spain for five years.
[Updated info: Tom currently teaches at Daytona Beach Community College, Florida. –Ed. May 2002]



Back to Text

Individual Test Results

Table 2. LATIN-BASED (N = 10)

Language Pre-test Post-test
Spanish 47 100
Spanish 70 93
Spanish 63 90
Spanish 60 90
Spanish 60 77
Spanish 73 83
Spanish 60 87
French 57 97
French 70 93
Portuguese 47 83
Mean 60.7 89.3
SD 8.94 6.98

Table 3. NON LATIN-BASED (N = 24)

Language Pre-test Post-test
Russian 20 100
Russian 47 97
Russian 40 93
Russian 70 93
Russian 57 90
Russian 40 63
Finnish 67 97
Estonian 30 83
Turkish 53 87
Vietnamese 20 93
Vietnamese 50 87
Vietnamese 30 80
Vietnamese 37 73
Chinese 47 83
Chinese 46 90
Japanese/Chinese 27 37*
Taiwanese 57 93
Taiwanese 37 93
Taiwanese 57 90
Taiwanese 43 77
Taiwanese 33 77
Korean 50 100
Korean 40 87
Korean 47 80
Mean 43.54 85.13
SD 13.17 13.61

* Without this student’s score, the mean & SD would be 44.26 and 12.98 respectively, for the pre-test, 87.22 and 9.16 for the post-test (n = 23).


© Copyright rests with authors. Please cite TESL-EJ appropriately.

Editor’s Note: Dashed numbers in square brackets indicate the end of each page for purposes of citation.