The modal system is a challenging structure for ESL/EFL learners (Celce-Murcia & Larsen-Freeman, 1999; Master, 1996). Difficulties experienced are attributable to the variety of forms and multiplicity of meanings used to express modality. Modality in English is typically expressed with modal auxiliaries, known simply as modals; examples include should, must, might, etc. These forms are divided into two categories: root (social obligation) and epistemic (logical possibility). Meanings associated with the root modalshould are those of obligation, a sense of what one is supposed to do. Epistemic meaning implies probability, what one expects ought to occur (Master, 1996).

The natural order hypothesis of Krashen’s (1982) monitor theory states that certain language features, such as modal forms, will be acquired before others regardless of what is taught. Consequently, manydevelopmental sequences(i.e., lists of ordered stages of development) have been proposed. However, preexisting developmental sequences for should have not been identified by the researchers.

This article reports the results of a pilot study conducted with one learner in her acquisition of the modal should. The following research questions were investigated:

1. How does the learner’s use of should during structured lessons differ from performance in unstructured conversation?
2. How does limited instruction impact use of should?

Ultimately, the researchers attempted to define a developmental sequence of the learner’s acquisition of should.

METHODOLOGY

The learner, Emi, is a Japanese housewife who has studied English with one of the researchers for 4 years, and has been studying English informally for 40 years following her high school English education. Emi has read many unedited versions of English language novels. She has a sizeable vocabulary; however, she lacks speaking fluency. For the study, Emi was asked what she would like to focus on. She chose modals. Prior to the study, Emi was already using the modals should and have to; however, she was aware of her errors with these forms and wanted to improve. This study involved 10 one-hour weekly tutoring sessions, each divided equally into two parts: free conversation and instruction. During free conversation, the researcher did not explicitly encourage modal use; however, questions such as “What should I do?” were used by the researcher to create opportunities for such usage.

The instructional parts of the first two sessions assessed Emi’s command of modals. Both researchers transcribed and verified all occurrences of should. An analysis was done using 76 obligatory occasions of should that occurred across all 10 sessions. Various ways Emi attempted to produce should were categorized and examined using frequency analysis. Explicit instruction on modals for stating rules (e.g., “You should do your homework”) was the focus of Sessions 3 and 4. In Sessions 5 through 10, modals for giving advice (e.g., “You should have a party for your parents”)were taught and practiced using instructional worksheets and role plays.

In Sessions 3 and 9, Emi studied the meanings of modals (should, can, must, and have to), the level of certainty each expresses, and the negated forms. Session 4 focused on modal sentence structure. Written homework was given in Sessions 3 through 8, allowing Emi to prepare utterances beforehand. These written utterances served as scaffolding during discussion-based lessons (Ellis, 2008). In Sessions 9 and 10, Emi practiced forming statements of rules and advice without advance writings.

RESULTS

Due to Emi’s frequent use of should, the study centered on this modal. Emi used should and have to without prompting during free conversation in Session 2, which shows some understanding of modal usage before structured lessons began. She also used modals without prompting in Sessions 6 and 7. However, during Sessions 5, 7, and 9, Emi missed opportunities to use modals during free conversation, even with prompting. It should be noted that the topic of conversation (e.g., giving advice about child-rearing) varied from session to session and could have contributed to the frequency of modal production.

During structured Sessions 4 to 8, Emi showed a remarkable ability to memorize and recite sentences prepared in her written homework. In Session 8, she was asked to recall the rules from her Session 3 homework and did so with only a few pauses and mistakes. However, there was a considerable drop in proficiency when she was asked to produce original statements in Sessions 5, 6, 9, and 10.

The obligatory occasion analysis of Table 1 shows a 43 percent increase in the correct usage of should.

Table 1. Obligatory Occasion Analysis of Should

Figure 1 is a frequency analysis of all uses of should, divided into four categories. Other includes five subcategories: a) should with auxiliary and main verb; b) should with inflected verb; c) negative should with verb; d) negative should without verb; and e) should with negative auxiliary verb and verb.

Figure 1 indicates most modals were produced during Sessions 5 through 8. Sessions 1 to 4 provided fewer opportunities to use modals due to assessment and instruction. The last two sessions involved longer role plays during which attention was not exclusively focused on modal production.

Figure 1. Frequency Analysis of All Should Utterances (Errors After Verb Ignored)

DISCUSSION

Unfortunately, there was not enough data from the unstructured portions of each session, as relatively few instances of should were produced. Despite expectations, the paucity of data made a meaningful comparison of Emi’s should utterances in unstructured conversations difficult.

A tentative developmental sequence of this learner’s acquisition of should is described below. The sequence was derived from the frequency analysis; other researchers have used a similar methodology for negation and interrogatives (Cancino, Rosansky, & Schumann, 1978). We were unable to find a well-established sequence of should in previous research. However, from our data, we identified the following sequence:

1. should + (missing V)
2. missing should + V
3. should + V

Emi stopped producing should utterances with a missing main verb by the sixth session. Thus, she seemingly progressed out of stage one. The remaining two stages are more difficult to define. At all times, Emi produced utterances in the form of should + V more frequently than the form “missing should + V” (i.e., not using should before a verb even though necessitated by the meaning and/or context of the utterance). However, in Sessions 5 through 10, Emi produced both forms. There was never a period in which she favored missing should + V phrases over should + V. Missing should + V was compared with should + V utterances: in Session 5, there were six more should + V utterances than missing should + V utterances; in Session 10, there were three more; in Session 9, there was only one more. This unexpected result is more striking considering that in Session 7, Emi left out should more often than she correctly produced should in clauses that were also correct. One might have expected the opposite. It appears that, even though Emi correctly used should more frequently by the end, she also omitted should (as in missing should + V utterances) quite frequently.

Just as Emi’s knowledge of basic sentence structure seemed to falter with the introduction of modals, production of correct should + V phrases may adversely affect remaining portions of her clauses. One explanation is that the increase in other errors is due to added cognitive demand about which modal to use (Cummins, 1982). Perhaps another is lack of planning time to attend to form and meaning aspects of the modal and other parts of the clauses (Skehan, 1998).

CONCLUSION

Clearly, more work is needed to achieve conclusive results. It was beyond the scope of this project to investigate modals other than should. Likewise, the effect of the topic of conversation on the frequency of modals production was not examined. Follow-up with Emi’s development of the modal should is planned. It will be interesting to see when she begins to consistently produce appropriate should + V structures. The developmental sequence outlined above cannot be generalized to other learners. However, it is possible that this sequence could serve as a starting point for continued study.

REFERENCES

Cancino, H., Rosansky, E., & Schumann, J. (1978). The acquisition of English negatives and interrogatives by native speakers. In E. Hatch (Ed.), Second language acquisition (pp. 207-230). Rowley, MA: Newbury House.

Celce-Murcia, M., & Larsen-Freeman, D. (1999). The grammar book. Boston, MA: Heinle, Cengage Learning.

Cummins, J. (1982). Tests, achievement, and bilingual students.Focus, 9, 2-9.

Ellis, R. (2008). The study of second language acquisition. Oxford, England: Oxford University Press.

Krashen, S. (1982). Principles and practice in second language acquisition. Oxford, England: Pergamon.

Master, P. (1996). Systems in English grammar: An introduction for language teachers. Upper Saddle River, NJ: Prentice Hall Regents.

Skehan, P. (1998). A cognitive approach to language learning. Oxford, England: Oxford University Press.