101017 s 0272263100011979

SSI A. 15, 225-243. Printed in the United Stales ol America.

E X P L I C I T I N S T R U C T I O N A N D I N P U T P R O C E S S I N G

Bill VanPatten The University of Illinois at Urbana-Champaign

Teresa Cadierno University ofAarhus

In this paper w e descr ibe a n exper iment in explicit instruction that c o m p a r e s traditional form-focused instruction a n d what w e call processing instruction. Traditional instruction involves explanat ion a n d output practice of a grammat ica l point. Processing instruction involves explanation a n d pract ice /exper ience processing input data , taking learner strategies in input processing as the starting point for determining what explicit instruction should look like. Pretest a n d posttest measures involving both a sentence-level interpretation (comprehension) task a n d a sentence-level production task w e r e submitted to a n analysis of var iance. Results reveal significant gains in both comprehension a n d product ion for subjects w h o exper ienced processing instruction. For those exper iencing traditional instruction, significant gains w e r e m a d e in product ion only.

Research on and discussion about the role of explicit instruction of grammar in second language (L2) classrooms has tended to focus on one of two issues. The first is whether or not instruction makes a difference and, by extension, whether or not explicit instruction is beneficial to second language acquisition (SLA) (e.g., Ellis, 1989; Krashen, 1982; Long, 1983; and others). The second issue is what can or should be taught. This includes the research on teachability-learnability (Pienemann, 1987), markedness (Eckman, Bell, & Nelson, 1988; Gass, 1982), parameter resetting (White, 1989), and the relationship of structure to meaning (Garrett, 1986). One issue that

We would like to thank all the participants at the Montreal Colloquium for a week of very useful discussion and comments on explicit grammar instruction. We are particularly indebted to Patsy Lightbown, Nina Spada, Lydia White, Stephen Krashen, Alice Omaggio Hadley, Jan Hulstijn, and the anonymous reviewers for SSLA who helped us to clarify our ideas. Errors in content rest with us.

c

1993 Cambridge University Press 0272-2631/93 SS.00 + .00 2 2 5

226 Bill V a n P a t t e n a n d T e r e s a C a d i e r n o

I II III input - in take - deve lop ing s y s t e m - output

Figure 1. Processes in second language acquisition.

has been overlooked in the research on explicit instruction is how grammar should be taught. That is, the nature of the instruction itself and the processes it attempts to modify in the learner have largely gone uninvestigated (see Tomasello & Herron, 1989, for a notable exception).

The purpose of the present paper is to examine the relationship between the nature of explicit instruction and its effects on SLA. Specifically, we examine the possible effects of two different types of instruction on the developing knowledge system of the L2 learner: instruction as the manipulation of output and instruction as structured or focused input processing. We will first situate input processing within a general model of SLA. We will then present the findings of an experiment that compares the relative effects of the two different kinds of instruction mentioned previously.

I N P U T P R O C E S S I N G I N S L A

SLA is multifaceted by nature, and more than likely no single theory of language or psychology will be able to capture how language acquisition happens (Ellis, 1990; McLaughlin, 1987; White, 1989). Nonetheless, as Long (1990) pointed out, any theory of SLA must acknowledge the role of comprehensible input in the development of the learner's internal grammar (see also Gass, 1988). Given the important role of comprehensible input in SLA and regardless of one's theoretical bent, SLA can be conceived of as sets of processes, as depicted in Figure 1 (from VanPatten, 1992).

Figure 1 attempts to capture three distinguishable sets of processes in acquisition. The first set of processes (I) converts input to intake. This is referred to as input processing and will be discussed below. From intake the learner must still develop an acquired system; that is, not all intake is automatically fed into the acquired system. The second set of processes (10 then includes those that promote the accommodation of intake and the restructuring of the developing linguistic system (McLaughlin, 1990; White, 1989). Finally, it is not clear from output studies that learner language is a direct reflection of acquired competence. Thus, a third set of processes (III) must be posited to account for certain aspects of language productionfor example, monitoring, accessing, control, and so on (Schmidt, 1992; Terrell,. 1991).

Input processing is concerned with those processes involved in Set I, the conversion of input to intake. While input processing can be examined from a variety of perspectives (Chaudron, 1985; Gass, 1988; Terrell, 1991; VanPatten, in press), we will use the notion of form-meaning connection to discuss the processes that are involved in the conversion of input to intake. In other words, the term input processing as it is used here involves those strategies and mechanisms that promote form-meaning connections during comprehension. (For a more detailed discussion of the

Input Processing 227

input - intake - developing system - output

focused practice

Figure 2. Traditional instruction in foreign language teaching.

input j intake - developing system output

processing mechanisms

1 focused practice

Figure 3. Processing instruction in foreign language teaching. relationship among form, meaning, and input processing, see VanPatten, in press.) As the learner processes an incoming input string, it must be tagged and coded in particular ways. If acquisition is to happen, the internal processors) must eventually attend to how the propositional content is encoded linguistically. For us, then, intake is that subset of the input that a learner comprehends and from which grammatical information can be made available to the developing system. Sharwood Smith (1986) distinguished between comprehension and acquisition vis-a-vis input processing. He suggested that for acquisition to occur the learner must attend to linguistic features in the input rather than only the message. Our notion of input processing and intake are consistent with this view.

I N S T R U C T E D S L A

To understand the connection that we would like to make between input processing and instruction, it is necessary to remember how traditional foreign language instruction in grammar occurs. Normally, this instruction focuses on the manipulation of learner output. In most foreign language classrooms, instruction occurs by explaining a grammatical concept and then having learners practice producing a given structure or form (see Figure 2). Given the rather important role that comprehensible input plays in SLA, the value of grammatical instruction as output practice is questionable if the intent of the instruction is to alter the nature of the developing system. Note that in Figures 1 and 2 the intake for the developing system is part of a chain of processes, and in the figures the arrows go from left to right and not right to left. It would seem reasonable, then, to suggest that rather than manipulate learner output to effect change in the developing system, instruction might seek to change the way that input is perceived and processed by the learner. This is depicted in Figure 3. Theoretically, altering input processing should have a significant impact

228 Bill VanPatten and Teresa Cadierno

on changing the internalized knowledge. (We will comment on the role of output in language teaching and learning in the Discussion section.)

To research such a hypothesis, one must first identify the strategies and mechanisms used by learners in getting intake from input. While not a great deal is known about input processing, there is some discussion in the literature about how learners process input (e.g., Chaudron, 1985; Gass, 1988; Hulstijn, 1989; Klein, 1986; Schmidt, 1990; VanPatten, 1990). In addition, research in child first language (LI) acquisition offers discussion of possible strategies used by 12 learners while processing input (e.g., Bever, 1970; Peters, 1985). One strategy that has received considerable attention uses word order to assign argument structure to an input string. Evidence from child LI and child and adult 12 studies shows that early- and intermediate-stage learners assign agent status to the first noun (phrase) of a string and object status to the second noun (phrase) (Bates et al., 1984; Bever, 1970; Ervin-Tripp, 1974; Gass, 1989; LoCoco, 1987; Nam, 1974; and others). While this may be a useful 12 strategy when applied to English input sentences (except for passives and clefts), it is not such a useful strategy for a language like Spanish. Spanish has flexible word order allowing a number of surface structure possibilities: SVO, SOV, OVS, and OV. Given that ambiguity can result, Spanish uses the case marker a to identify objects when more than one noun is capable of performing the action: El senor sigue a la senora/ A la senora la sigue el hombre ("The man follows the woman"), or El gato sigue al perro/Al perro lo sigue el gato ("The cat follows the dog").

With clitic object pronouns, word order is less flexible, resulting in an obligatory preverbal position for object pronouns when the verb is a simple finite verb. The subject may be placed before or after the verb, depending on features of syntax, discourse, style, and pragmatics. Either of the following sentences means 'The man follows her":

El serior to sigue. The man-SUBJ her-OBJ foUows. La sigue el serior. Her-OBJ follows the man-SUBJ.

Research on learners of Spanish has shown that input strings in which subject-object (agent-object) order is reversed are misassigned argument structure (e.g., Lee, 1987; LoCoco, 1987; VanPatten, 1984):

La sigue el senor. 'She follows the man.

The result in the learner's developing system as revealed by examination of learner output is an absence of object pronouns or the misuse of object pronouns as subjects of a sentence, incorrect placement of object pronouns, the (resetting of an incorrect


parameter (i.e., the learner may assume that Spanish is [-null-subject]), the absence of the case marker a, and difficulty in the acquisition of a certain class of verbs that obligatorily place subjects in postverbal position. In short, learners fail to see that Spanish is not a rigid SVO language. We make no claims about the relevance of this particular processing strategy for the acquisition of English or other languages. In the present study we will focus on the acquisition of Spanish by English speakers.

T H E P R E S E N T S T U D Y

In VanPatten and Cadierno fin press), we examine the outcome of explicit instruction in processing input. Unlike traditional explicit instruction in which output is manipulated, we sought to alter the strategy by which learners make form-meaning connections when exposed to input strings (cf. Figures 2 and 3). In the present study we expand our previous research by replicating the research and including more subjects per experimental cell. We once again research the impact of attempting to alter learners' processing of input containing non-SVO order and we compare three groups: one that received traditional foreign language instruction in object pronouns, one that received processing instruction in the same, and another that received no explicit instruction at all regarding object pronouns. We asked the following questions at the outset of the study:

1. Does altering the way in which learners process input have an effect on their developing systems?

2. If there is an effect, is it limited solely to processing more input or does instruction in input processing also have an effect on output?

3. If there is an effect, is it the same effect that traditional instruction has (assuming an effect for the latter)?

S u b j e c t s

Six second-year university-level Spanish classes at the University of Illinois were selected for inclusion in the present study (/V = 129). Two classes were randomly assigned to each one of the three treatment groups: no instruction, processing instruction, and traditional instruction. Both pretesting and recurrent absenteeism from posttesting phases (see later) eliminated subjects in each group. The final number of subjects was 80, distributed in the following manner: no instruction [N = 27); processing instruction (/V = 27), and traditional instruction (/V = 26).

Subjects were enrolled in a program with a communicative methodology modeled on the Natural Approach, where emphasis is placed on developing communicative skills in the L2. Gasses met 4 days a week with the bulk of class time spent on interaction, listening, and reading. Grammar instruction was limited to textbook and workbook exercises done as homework. However, per their regular syllabus, subjects were not scheduled to receive explicit instruction in object pronouns and word order during the time period of this investigation.


I n s t r u c t i o n a l P a c k e t s

Two packets were constructed for use during the instructional treatment. These packets reflected two different approaches to the teaching of the clitic direct object pronouns. The first approach consisted of traditional grammar teaching and output practice, while the second, called processing instruction, involved teaching the subjects to process input sentences differently from the strategies described earlier.

Traditional instruction involved presenting the subjects with explanations concerning the form and position of direct object pronouns within the sentence and then giving them practice in how to make sentences with those pronouns. Presentation involved a paradigmatic chart that included all persons (both singular and plural) and an explanation of what objects and object pronouns are. The presentation also included a description with examples of correct placement of object pronouns within a sentence. Subjects were taught that object pronouns always preceded simple conjugated verbs but could be placed at the ends of infinitives and present participles.

In terms of practice, traditional instruction involved moving the subjects from mechanical form-oriented practice (oral and written transformation and substitution drills) to meaningful practice (oral and written questions, simple sentence formation), and, finally, to more open-ended communicative practice (oral and written question and answer, conversation). At all times the traditional instruction focused the learners on producing the targeted items. This practice included both oral and written production. An example of an oral practice follows (English translations provided here):

With another student, make up and answer questions following the model.

Model: comer en lu casa ("to eat at your house") - iCudndo me invitas a comer en tu casa? "When are you going to invite me to your house?" Te invito para el sibado. "I'll invite you over for Saturday."

1. cenar en tu casa ("to dine at your house") 2. almorzar "to lunch" land so on]

An example of a written activity used in traditional instruction is the following (English translations provided here):

Describe what is happening in each of the following drawings. Use object pronouns in your description.

1. i Qui hace la mujer? ("What is the woman doing?") (accompanied by a drawing ol a woman grabbing a man and then kissing him)

2. iQui hace el hombre? ("What is the man doing?") (accompanied by a drawing of a man meeting two women and then inviting them to the movies)

[and so on]


The instructional packet used for this type of instruction was based on a best-selling first-year college-level textbook and its workbook (see Knorre et al., 1985). Other examples of the types of activities used in traditional instruction are presented in Appendix A.

Processing instruction, on the other hand, involved (a) teaching the subjects how to interpret OVS strings correctly, both when the 0 consisted of a full noun and when it consisted of a clitic object pronoun, and (b) having the students respond to the informational content of OV strings. The instructional packet used for this type of instruction was based on a new first-year college-level text that incorporates this type of instruction (see VanPatten, Lee, Ballman, & Dvorak, 1992).

In processing instruction the presentation of the direct object pronouns first contrasted the grammatical concepts of object and subject of a verb and then presented both subject and object pronouns. This presentation was followed by explanations of important points to keep in mind about the position of object pronouns in Spanish. It was emphasized that in Spanish, unlike English, it is possible to have sequences of the type object pronoun-verb-subject (see previous examples). Subjects in the processing group also learned that full object nouns may be placed before the verb and these nouns are case marked with the preposition a:

A Maria la llama Juan. Mary-OBJ her-OBJpro calls John-SUBJ. "John calls Mary."

Two types of activities followed the presentation and explanations of the object pronouns. One type had subjects listening to utterances or reading sentences and then somehow demonstrating that they had correctly assigned argument structure to the string. This typically involved selecting the drawing that best represented what was heard/read or selecting the best English rendering of the Spanish sentence. The second type of activity had subjects respond to the content of an utterance or sentence by checking "agree" or "disagree," "true for me" or "not true for me," and so on. In several activities subjects read a very short passage in which, subsequent to reading the passage, sentences with object pronouns were highlighted and subjects were asked what those particular utterances meant. Activities used both aural and written stimuli, although a majority were written. An example of an aural activity for processing instruction follows:

Listen as your instructor reads a sentence. Select the best interpretation from the English renderings.

1. a. My parents call me. b. I call my parents.

(Instructor reads aloud: Me llaman los padres.) [and so on]

An example of a written activity used in processing instruction follows (English translations provided here):


Indicate whether or not each statement about your parents applies to you. Then share your responses with a classmate.

Si, se me aplica. No, no se me optica. 1

2

(and so on]

Other examples of activities used in processing instruction are presented in Appendix B. What should be underscored here is that at no point did processing instruction involve the production of the pronoun forms by the learners.

Vocabulary, which consisted of frequent items, was the same for both types of instruction. The amount of practice that learners received (i.e., the number of times one produced a sentence in the traditional group vs. the number of times one interpreted or responded to a sentence in the processing group) was also roughly the same for both instructional treatments.

To summarize this section, we highlight the following. First, both experimental groups received explanation about object pronouns and where to place them in the sentence. Both groups received the same amount of practice and both utilized the same vocabulary. The instruction differed in two ways: (a) The processing group was told that language learners often misinterpret NVN sequences, and (b) the traditional group did not practice interpreting sentences, and the processing group did not practice producing sentences. In short, the fundamental difference between the two treatments involves the type of practice the students received.

P r e t e s t a n d P o s t t e s t s

A pretest/posttest procedure using a split-block design (see below) was used to assess the effect of instruction. All tests (i.e., the pretest and the three posttests) consisted of both interpretation tasks and written production tasks. In between the interpretation and production tasks, a distractor task consisting of writing answers to questions unrelated to and not containing the grammatical item under study was given to the subjects.

The interpretation tasks for all tests consisted of 15 aural sentences. Five of these sentences consisted of SVO word order sentences, which served as distractors. Of the remaining 10 test sentences, five consisted of sentences of the following type:

Alchico to saluda lachica. Theboy-OBJ him-OBJpro greets the girl-SUBJ. 'The girl greets the boy."

The other five sentences were of the following type:

Los ttamo con frecuencia por let-ifono. ("I call them on the phone frequently.") Los visito los fines de semana. ("I visit them on the weekends.")


Lo saluda la chica. Him-OBJ greets the girl-SUBJ. "The girl greets him."

For the interpretation tasks, subjects were asked to match each sentence they heard with one of two pictures that were simultaneously presented to them via overhead projector. The two pictures represented the same action, the difference between them being who the agent was and who the patient of the verb was. For example, for the sentence Lo saluda la chica, subjects had to choose between a picture of a boy waving to a girl (who did not wave back) and a picture of a girl waving to a boy (who did not wave back).

The production section of the test was a simple written task based on activities used in traditional instruction. It included five items, each of which consisted of an incomplete sentence. The subjects' task was to complete the sentence according to a visual clue. The test items were constructed along the following lines:

El chico piensa en la chica y entonces .'The boy is thinking about the girl and then ."

Each item was accompanied by two drawings that depicted the content of the sentence. In the preceding example, the first picture showed a boy sitting at home thinking about a girl, and the second showed the boy calling the girl on the phone. It was assumed that the second picture would guide the subject to create a sentence in which an object pronoun should be used. In the case of the preceding example, the subject should have completed the sentence in the following manner: "The boy is thinking about the girl and then he calls her." While the visual cues for the interpretation task were projected on a screen via overhead projector, the drawings used for the production task were on the page along with each item. All test items for both task types involved simple present tense sentences.

I n s t r u c t i o n a l a n d D a t a C o l l e c t i o n P r o c e d u r e s

A split block design was used that included four versions of the same test: A, B, C, and D. Half of the classes received version A as the pretest and half received version B. To ensure that subjects understood the nature of the interpretation task, they were given a practice item. To control for familiarity of vocabulary, a list of Spanish-English equivalencies was provided to subjects prior to testing. Subjects were given 2 min to study the list and familiarize themselves with any new words. The interpretation task was always administered before the production task.

After the pretest was administered, classes were randomly assigned to one of three instructional treatments: traditional instruction, processing instruction, and no instruction, the last being the control group. Subjects in the first two instructional treatments received instruction during classtime in lieu of their normal routine. For both groups explicit instruction consisted of two consecutive class days with no homework.


It is difficult to control for instructor bias in studies such as the present one. Nonetheless, certain measures were taken, the first being that all experimental instruction was performed by the same person. At the time of the investigation, she was an experienced teaching assistant pursuing a doctoral degree. She was not the subjects' regular classroom instructor and she did not instruct them other than on the 2 days during which the instructional treatment was carried out. At the outset of the study, she had no knowledge of processing instruction but was well versed in traditional instruction. What is important to note in terms of bias is that when asked to cooperate in this study, she formulated her own hypotheses about the outcome. She believed that the traditional group would make gains in and be better than the other two groups in producing object pronouns. She also believed that the processing group would make gains in and be better than the other two groups in comprehending utterances with object pronouns. It was not until the first round of posttest results were analyzed that the two researchers sat and discussed the hypotheses. Only then was she led to reconsider her hypotheses.

At the end of the 2nd day of instruction, subjects were given the first posttest. Those subjects who received version A as the pretest received version B as the first posttest. Those who received version B as the pretest received version A as the first posttest. Versions C and D were administered 1 week after instruction and again 1 month later. As in the case of the pretest and the first posttest, those subjects who received version C as the second posttest received version D as the third posttest. Those who received version D as the second posttest received version C as the third posttest. Subjects must have been present at all three posttests (in addition to the pretest) to be included in the study (see earlier; final N = 80). %

It should be noted that the no instruction (control) group had regular class hours and did not receive any special instruction during the experimental phase. Their instructor continued with normal topics and activities outlined in the course syllabus and was unaware of the experimental instruction occurring with the other two groups. She was under the impression that we were simply gathering data. She was present for the ptetest but was not present for the posttests. For all three groups, experimentation and testing took place in the subjects' regular classrooms during their regular class hours.

S c o r i n g P r o c e d u r e s

Raw scores for each subject were calculated for both the interpretation task and the production task. With respect to the interpretation task, each correct response to the 10 test items was given a score of 1 for a possible total of 10. Correct responses consisted of correctly matching the sentence heard to a drawing seen. Incorrect responses received a score of 0. Because the impact of instruction in a study of this kind is measured by an increase in knowledge, we decided to eliminate certain subjects from the beginning if they demonstrated a tendency to not use the word order strategy on the pretest. An arbitrary score of 8 out of 10 on the interpretation task was set so that any subject with 8 or above on the pretest task for interpretation was eliminated from the study. (Again, per the description of the subjects given


previously, after pretesting and eliminating any subjects who missed a posttest, final N = 80.)

For the production task, raw scores were calculated by counting each correct response to the five test items. Responses were given a score of 2 points if the subjects produced a correct direct object clitic pronoun form in the correct position within the sentence. Responses scored 0 points if the subjects produced no clitic pronoun form at all (even if the sentence was correct by all other standards). For all those cases in between, a liberal scoring procedure was adopted. Responses were given a score of 1 point if the subjects produced either a correct direct object clitic pronoun form in the incorrect position, an incorrect clitic pronoun form in a correct position, or an incorrect clitic pronoun form in an incorrect position. The 2, 1, 0 scoring procedure was used because an either/or scoring procedure would not reveal possible intermediate effects of instruction. In other words, instruction could have had an impact on the learner's developing system but not necessarily resulting in nativelike accuracy or use. Because there were five items with a possible score of two each, the production task was also worth 10 points. As in the interpretation task, subjects were eliminated from the study if they scored 8 points or above on the pretest.

A n a l y s i s

Raw scores were submitted to two separate one-way analyses of variance (ANOVA) with a repeated measures design. The first ANOVA was conducted on the interpretation data and the second on the production data.

R E S U L T S

An ANOVA conducted on the pretests alone revealed no differences between the groups before instruction (o = .30 for the interpretation task and p = .26 for the production task). We are thus confident that any comparative effects due to instruction are not related to prior knowledge or ability of any one group.

The ANOVA with repeated measures conducted on the raw scores of the interpretation tasks revealed a significant main effect for instruction (df = 2, F = 33.129, p = .0001), a significant main effect for test (pretest vs. posttest, df=3,F= 76.921, p = .0001), and a significant interaction between instruction and test (df = 6, F = 21.913, p = .0001). The results of this ANOVA are, displayed graphically in Figure 4, and the means and standard deviations of each group's scores are displayed in Table 1. A posthoc Sheffe test revealed that the effect for instruction was due to the following contrasts: processing better than no instruction (p = .0001), processing better than traditional instruction (p = .0001), and no significant difference between traditional and no instruction (p = .12). In short, processing instruction was superior to the other two instructional types vis-a-vis the interpretation test. The significant main effect for test was subjected to a means comparison, which revealed that the main effect was due to posttests being significantly different from pretests but none of the posttests being significantly different from each other. No posthoc test was


none processing traditional

Figure 4 . Results of ANOVA with repeated measures using raw scores on the interpretation test.

Table 1 . Mean raw scores and standard deviations for interpretation task

No Instruction Traditional Processing ( I V = 27) ( I V = 26) (JV = 27)

M SD M SD M SD

Pretest 1.11 1.05 1.31 1.64 1.74 1.58 Posttest 1 1.41 2.41 3.04 3.22 8.07 2.11 Posttest 2 2.26 2.77 3.42 3.11 7.19 2.45 Posttest 3 2.22 2.14 3.89 3.23 7.41 2.24

available on the interaction. In light of the results given in Figure 4, it would seem that the interaction is due to the processing group being superior to the other groups on the posttests.

The answer to our first research question is yes, processing instruction has some effect on how learners process input, which in turn has an effect on the developing system of the learners. In addition, we can partially answer the third research question. It appears that there is some differential effect for processing instruction since traditional instruction apparently did little to improve learners' success in making form-meaning connections while processing input. To answer our second question and to finish answering the third, we now turn our attention to the results of the production task.

The ANOVA with repeated measures conducted on the raw scores of the production tasks also yielded a significant main effect for instruction (df = 2, F = 25.214, p = .0001), a significant main effect for test (pretest vs. posttest, df=3,F= 97.016, p = .0001), and a significant interaction between instruction and test (df = 6,

1 0 pretest


l~l pretest ffl posttest 1 I posttest 2 posttest 3

none processing traditional

Figure 5 . Results of ANOVA with repeated measures using raw scores on the production test.

Table 2 . Mean raw scores and standard deviations for production task

No Instruction Traditional Processing ( tV = 27) f/v = 26) ( IV = 27)

M SD M SD M SD

Pretest 1.59 2.29 2.69 2.60 2.19 2.37 Posttest 1 3.41 3.44 8.54 2.52 8.89 2.47 Posttest 2 3.41 3.85 9.04 2.25 8.19 3.00 Posttest 3 4.59 3.83 8.12 2.72 8.11 2.86

F = 8.708, p = .0001). The results of this second ANOVA are displayed graphically in Figure 5, and the means and standard deviations of each group's scores appear in Table 2. A posthoc Shefte test revealed the following contrasts: traditional better than no instruction (p = .0001), and processing better than no instruction (p = .0001). TTiere was no significant difference between processing instruction and traditional instruction. In short, traditional instruction was not superior to processing instruction on the production task. A posthoc Means Comparison conducted on the main effect for test yielded the same results as for the interpretation task. Again, no posthoc was available for the interaction, but from Figure 5 it can be assumed that the interaction is due to processing and traditional grouping together in the posttests compared to no instruction.

To continue answering our research questions, the answer to the second question is also yes, processing instruction does have an effect on production, at least in the way that production was tested here. Given the answer to this question, the answer to the third research question is that processing and traditional instruction appar-


ently do not impact the same way on the learner. These results parallel those in VanPatten and Cadierno fin press). What appears to have happened in this study is that processing instruction altered the way in which the subjects processed input, which in turn had an elfect on the developing system and what the subjects could access for production. The converse was not the case for the subjects in the traditional group. That is, while traditional instruction apparently had an impact on what the subjects could access for production, it had little impact on how the subjects processed input. While the traditional group did improve somewhat over time, their scores were not significantly different from those of the no instruction group. We therefore conclude that the slight gains in interpretation made over time by the subjects in the traditional group cannot be attributed to instruction. We now turn to some general discussion.

D I S C U S S I O N

The results reported in the previous section are interesting. While the processes depicted in Figure 3 help us to understand why processing instruction would have an effect on production, we see only two possible explanations why traditional instruction with output had little effect on interpretation. The first is that the subjects in the production group learned to perform the task but did not acquire any new language. However, this is problematic in that to perform a language task, one must have some kind of knowledge. To produce a sentence, even in a mechanical sense, one must draw upon some knowledge source to put together that sentence. In addition, the subjects in the processing group were able to perform a production task that was not part of their instruction. We therefore believe that the second explanation is more tenable: that traditional grammar presentation and practice do not enhance how learners process input and therefore do not provide intake for the developing system. Instead, traditional instruction results in a different knowledge system. Krashen (1982 and elsewhere) has suggested that learners may develop two systemsan acquired competence and a learned competenceand has claimed that traditional instruction results in learned competence. He argues that only by getting comprehensible input can the learner's internal system develop. Likewise, Schwartz (this issue) has suggested that the language module in the mind can only operate on primary linguistic data (input) and that explicit practice and negative evidence are not usable by the module. Explicit practice and negative evidence can result in what she calls learned linguistic knowledge, or LLK, but this is a distinct system from the underlying competence used by the language module. While Krashen and Schwartz may or may not agree with a focus on input as a teaching strategy, the data in our study would appear to support their claims about the type of knowledge source resulting from traditional instruction.

We now turn our attention to some of the methodological objections that could be raised regarding the study. First, while we may have shown that instruction in processing input transfers to production under controlled conditions, we have not shown that it transfers to production under more spontaneous circumstances. We agree with this objection but also offer the following argument. We did not set out


to test whether or not instruction results in better communicative performance. We set out to ascertain whether or not instruction involving a focus on input processing resulted in similar or different effects compared to traditional instruction. We believe we have done this. As part of a series of studies on this question, we will turn our attention in a future study to samples of spontaneous speech and writing gathered before and after instruction to see whether or not an effect of the two instructional types on more spontaneous performance can be found.

Another methodological objection to this study (and others like it) is that its essentially quantitative nature obscures possible qualitative differences in the production task. Given the scoring procedure of 2 points for a well-formed sentence, 1 point for a variety of nonnativelike sentences, and 0 points for sentences that basically reveal no learning at all, the question could be raised whether or not the nonsignificant difference between the traditional and processing groups masks different types of output errors or production strategies. For example, one group could have received a balanced number of Os and 2s while the other group received mostly Is. We addressed this question by conducting a multiple ANOVA using the number of 2s, Is, and Os received by each subject on the production task as the dependent measure and instruction as the independent variable. The results did not yield any main effect and a probing with a posthoc Sheffe test revealed that no differences existed between any of the possible group comparisons. In addition, we looked at the production sentences to see whether or not the sentences scored as 1 were qualitatively different between the two groups. We could not discern any difference.

A final comment that we would like to make concerns the types of linguistic items used in this study. In the present study, we used what can be considered a salient linguistic item: free-standing object pronouns and word order. But what of bound morphology? In Cadierno (1992) processing and traditional instruction are compared using past tense verb morphology as the linguistic item. Her results reveal the same pattern of development as in the present study: Processing instruction has a significant effect on how learners perceive and produce past tense forms in Spanish; traditional instruction has an effect only on production of past tense forms.

Before concluding, it is worth clarifying our position on the role of output in both learning and teaching. First, we do not advocate abandoning communicative activities and tasks that provide opportunities for making output. Given the set of processes labeled Set III in Figure 1, it is clear that learners need to develop their abilities in accessing the developing system for fluent and accurate production. We see these types of output activities following instruction that focuses on input. In other words, explicit grammar instruction should first seek to make changes in the developing system via a focus on input and only afterward should instruction provide opportunities for developing productive abilities. Second, learners can and often do use each others' output in the classroom as input. Thus, we feel that, in addition to the fluid and "freer" interaction that often happens in communicative classrooms, it is important for instructors to also develop focused output activities that encourage learners to be accurate while also attending to meaning. We are not advocating mechanical drilling or what Savignon (1983) has called "forms without


function" (p. 17). Instead, we see that along with input and free-flowing communication in the classroom, there is room for tasks that are structured around particular grammatical points where real messages are communicated and learners attend to both content and form.

C O N C L U S I O N

We believe to have made a strong connection between input processing and instruction in the present paper. In light of the claims made for the importance of comprehensible input in SLA, instruction as direct intervention on learners' strategies in input processing should have a significant effect on the learner's developing system. The results of the present study suggest that this is so. The present study, along with VanPatten and Cadierno (in press) and Cadierno (1992), demonstrates that instruction is apparently more beneficial when it is directed at how learners perceive and process input rather than when it is focused on practice via output. Learners who receive instruction that attempts to alter input processing receive a double bonus: better processing of input and knowledge that is apparently also available for production. The results are important, then, not only because of what they might contribute to the ongoing discussion of the effects of instruction but also for the support that they give to input processing as a critical aspect of SLA.

REFERENCES

Bates, E., MacWhinney, B., Caselli, C, Devescovi, A., Natale, F., & Venza, V. (1984). A cross-linguistic study of the development of sentence interpretation strategies. Child Development, 55,341-354.

Bever, T. G. (19701 The cognitive basis for linguistic structures. In J. R. Hayes (Ed), Cognition and the development of language Ipp. 279-362). New York: Wiley.

Cadierno, T. (1992) Explicit instruction in grammar: A comparison of input based and output based instruction in second language acquisition. Unpublished doctoral dissertation, The University of Illinois at Urbana-Champaign.

Chaudron, C (1985). Intake: On models and methods for discovering learners' processing of input. Studies in Second Language Acquisition, 7,1-14.

Eckman, F., Bell, L , & Nelson, 0. (1988). On the generalization of relative clause instruction in the acquisition of English as second language. Applied Linguistics, 9,1-11.

Ellis, R. (1989). Are classroom and naturalistic acquisition the same? A study of the classroom acquisition of German word order rules. Studies in Second Language Acquisition, 11,305-328.

Ellis, R. (1990). Instructed second language acquisition. Cambridge, MA: Blackwell. Ervin-Tripp, S. M. (1974) Is second language learning like the first? TESOL Quarterly, 8,111-127. Garrett, N. (1986). The problem with grammar What kind can the language learner use? 77te Modern

Language Journal, 70,133-148. Gass, S. M. (19821 From theory to practice. In M. Hines & B. Rutherford (Eds.), On TESOL '81 Ipp. 129-139)

Washington, DO TESOL Gass, S. M. (19881 Integrating research areas: A framework for second language studies. Applied Linguistics,

9,198-217. Gass, S. M. (1989) How do learners resolve linguistic conflicts? In S. M. Gass & J. Schachter (Eds.), Linguistic

perspectives on second language acquisition ipp. 183-199). Cambridge: Cambridge University Press. Hulstijn, J. H. (1989) Implicit and incidental second language learning: Experiments in the processing of

natural and partly artificial input. In H. W. Dechert & M. Raupach (Eds.), Interlingual processing Ipp. 49-73). Tubingen: Narr.

Klein, W. (1986). Second language acquisition. Cambridge: Cambridge University Press. Knorre, M., Oorwick, T., Higgs, T., VanPatten, B., Ferran, F., 4 Lusetti, W. (1985). Pantos de partida. New

York: Random House. Krashen, S. D. (1982). Principles and practice in second language acquisition. Oxford: Pergamon Press.


Lee. J. F. (1987). Morphological factors influencing pronominal reference assignment by learners of Spanish. In T. A. Morgan, J. F. Lee, 4 B. VanPatten (Eds.), Language and language use: Studies in Spanish Ipp. 221-232). Lanham, MD: University Press of America.

LoCoco, V. (1987). Learner comprehension of oral and written sentences in German and Spanish: The importance of word order. In B. VanPatten, T. Dvorak, 4 J. F. Lee (Eds.), Foreign language learning: A research perspective (pp. 119-129). Rowley, MA: Newbury House.

Long, M. H. (1983). Does second language instruction make a difference? A review of research. TESOL Quarterly, IT. 359-382.

Long, M. H. (1990). The least a second language acquisition theory needs to explain. TESOL Quarterly, 24, 649-666.

McLaughlin, B. (1987) Theories of second language acquisition. Philadelphia: John Benjamins. McLaughlin, B. (1990) Restructuring. Applied Linguistics, II, 113-128. Nam, F- (1974, March). Child and adult perceptual strategies in second language acquisition. Paper presented

at the TESOL Annual Meeting, Los Angeles. Peters, A. (1985). Language segmentation: Operating principles for the perception and analysis of language.

In D. I. Slobin (Ed.), The cross-linguistic study of language acquisition (pp. 1029-10671 Hillsdale, NJ: Erlbaum.

Pienemann, M. (1987). Psychological constraints on the teachability of languages. In C Pfaff (Ed.), First and second language acquisition processes (pp. 143-168). Cambridge, MA: Newbury House.

Savignon, S. (1983). Communicative competence: Theory and classroom practice. Reading, MA: Addison-Wesley.

Schmidt, R. (1990). The role of consciousness in second language learning. Applied Linguistics, II, 129-158. Schmidt, R. (1992) Psychological mechanisms underlying second language fluency. Studies in Second Lan

guage Acquisition, 14,357-385. Schwartz, B. (19931 On explicit and negative data effecting and affecting competence and linguistic behavior.

Studies in Second Language Acquisition, 15,147-163. Sharwood Smith, M. (1986). Comprehension vs. acquisition: Two ways of processing input. Applied Linguis

tics, 7,239-256. Terrell, T. (1991). The role of grammar instruction in a communicative approach. Modem Language Journal,

75,52-63. Tomasello, M., 4 Herron, C. (1989). Feedback for language transfer errors: The garden path technique.

Studies in Second Language Acquisition, 11,385-395. VanPatten, B. (1984). Learners' comprehension of clitic pronouns: More evidence for a word order strategy.

Hispanic Linguistics, 1,57-67. VanPatten, B. (1990). Attending to form and content in the input. Studies in Second Language Acquisition,

12,287-301. VanPatten, B. (1992). Second language acquisition and foreign language teaching: Part 2.77ie ADFL Bulletin,

23,23-27. VanPatten, B. fin press). Input processing and second language acquisition: On the relationship between

form and meaning. Festschrift for Tracy David Terrell. New York: McGraw-Hill. VanPatten, B., 4 Cadierno, T. (in press). Second language acquisition as input processing: A role for instruc

tion. The Modem Language Journal VanPatten, B., Lee, J. F., Ballman, T. L . 4 Dvorak, T. (1992). iSabias que...? New York: McGraw-Hill. White, L. (1989). Universal Grammar and second language acquisition. Philadelphia: John Benjamins.


APPENDIX A

S A M P L E A C T I V I T I E S U S E D I N T R A D I T I O N A L I N S T R U C T I O N

A. Directions: Imagine that you are in the following situations, performing the indicated tasks. A friend asks you about particular items. Answer logically. Follow the models.

I. Vd. estd haciendo la maleta para un viaje a Acapulco. ("You are packing for a trip to Acapulco.") lEl traje de bono? ;Ctaro que lo necesito! (The swimming suit? Of course I need it") ArUculos: las sandalias, las galas de sol, los parttalones cortos, las camisetas, etc ("Articles: sandals, sunglasses, shorts, T-shirts, etc").

B. Directions: Rephrase sentences, changing direct object nouns to pronouns as needed.

1. El camarero trae los vasos y pone los vasos en la mesa. (The waiter brings the glasses and puts the glasses on the table.") -f camarero trae los oasos y los pone en la mesa.

(


APPENDIX B

S A M P L E A C T I V I T I E S U S E D I N P R O C E S S I N G I N S T R U C T I O N

Actividad B. Listen to the speaker on the tape. Match each sentence you hear with one of the statements below.

1. A man is calling me. I am calling a man.

2. My parents visit me. I visit my parents.

3.1 am pleasing to my family. My family is pleasing to me.

4. We are greeting a friend. A friend greets us.

5. Our relatives don't understand us. We don't understand our relatives.

Actividad Each sentence corresponds to something that you might do to your parents. Check which ones apply to you. Compare your responses with a classmate.

1. los llamo con frecuencia por telffono. 2. Los visito los tines de semana. 3. los visito por lo menos una vez at mes. 4. Los abrazo cuando los veo. (abrazar = to hug) 5. Los comprendo muy bien. 6. Los ignoro comptetamente. 7.ios ? .

Did you notice that there are no explicit subject nouns or subject pronouns in each sentence? Since the yo form of the verb can only refer to yo, no subject pronoun is needed. All of the above sentences are of the simple word order object pronoun + verb.

101017 s 0272263100011979

Documents

t t e n

n d t e r e s

d i e r n o

explicit grammar instruction

nature of explicit instruction

i c i t i n s t r u

n d product ion

n d output practice