Second language acquisition (SLA) research has focused substantially on the role that input plays in language th learning since the final decades of the 20 century. Undoubtedly, input is fundamental to the acquisition process of a second language (L2) (e.g., Carroll, 2001; Chaudron, 1985; Ellis, 1994; 1998a; Gass & Madden, 1985; Krashen, 1981, 1982, 1985, 1989; Larsen-Freeman & Long, 1991; Schwartz, 1993; Swain, 1985; VanPatten, 1994, 1996, 2002, 2004; White, 1989). It is generally assumed that SLA is somehow input dependent (Gass, 1997; VanPatten, 2004). VanPatten (2004), in particular, posits that “at some level input is the primary initial ingredient for the development of competence [in L2], however one construes that competence” (p. 35).

The term input in the context of SLA commonly refers to “language data that the learner is exposed to, that is, the learner's experience of the target language in all its various manifestations....[or] the 'language bath'” (Sharwood Smith, 1993, p. 166). Print, audio and visual materials, interactions with speakers in L2, and the L2 classroom make potential sources of input among a whole host of others. On top of that, Krashen's (1985, 1989) Input Hypothesis, “the most influential theory of the role of input” (Young, 1989, p. 123), postulates that “more comprehensible input [i.e., language adapted roughly to the comprehension level of the learner] clearly results in more language acquisition” (Krashen, 1989, p. 411, emphasis added). Some modification of input is also seen as necessary or useful in order to provide comprehensible input to the learner (e.g., Day, 1986; Gass & Madden, 1985; Long, 1983, as cited in Gregg, 2001). For instance, Hatch (1983) and Larsen- Freeman (1985) characterize a type of input beneficial to language learners, especially beginners and/or children. The L2 input in communicative situations is simplified or in some way “reduced” in terms of rate of speech, vocabulary, syntax, discourse, speech setting and others (as cited in Lee & VanPatten, 1995). In addition, Long (1983, as cited in Ellis, 1993, p.55) recognizes the distinct advantage of discourse modifications as spontaneous or naturally simplified input, such as “clarification requests, confirmation checks, comprehension checks, selfrepetitions and other repetitions” for negotiation of meaning in terms of L2 development (over and as opposed to pre-planned or pedagogically simplified input). VanPatten and Cadierno (1993b), furthermore, confine input in L2 acquisition to “language that encodes meaning…to which the learner attends for its propositional content [meaning]” (p. 46). VanPatten (1996) clarifies this construct.:

Although researchers hold different perspectives on [SLA] and may use different frameworks with which they investigate factors affecting [SLA], all concur in that meaning-bearing input is essential to [SLA]. What this means is that learners must be exposed to samples of language (and in great amounts) that are used to communicate information…. Without meaningbearing input learners cannot build a mental representation of the grammar that must eventually underlie their use of language (p. 5).

It must be understood, however, that not all input becomes intake, a term first coined by Corder (1967). Learners do not take in all available data of the language bath. Hence, one major psycholinguistic inquiry that interests researchers is what the learners perceive in the input for prospective cognitive processing. How that takes place is equally yet another (Sharwood Smith, 1993). VanPatten explains:

We know, however, that learners process input as they attempt to comprehend the message(s) contained in it. Processing the input involves “filtering” it in various ways. What learners actually wind up with after processing the input is a reduced, sometimes slightly altered set of data that theorists call intake. The brain uses intake, and not raw input data, to create a linguistic system (Lee & VanPatten, 1995, p. 94, emphasis original). [That] subset of filtered input that the learner actually processes and holds in working memory during on-line comprehension contains grammatical information as it relates to meaning that learners have comprehended (or think they have comprehended) (VanPatten, 2002, p. 761).

To put it simply, intake is that language which, according to Slobin (1979, as cited in Shook, 1994), is “extracted” or “segmented” (p. 58) or, what Tomlin and Villa (1994) refer to as, “detected” (p. 192) by the learners.

More interestingly, an increasing body of SLA research demonstrates that intake is made possible most likely when attention is given to input (e.g., Alanen, 1995; Leow, 1997, 1998a, 1998b; Long 1991; Schmidt, 1990, 1993, 1994, 1995, 2001; Tomlin & Villa, 1994; VanPatten & Cadierno, 1993a, 1993b; White, 1998; Wong, 2001). Schmidt (2001), for example, claims that “attended learning [which begins with noticing] is far superior, and for all practical purposes, attention is necessary for all aspects of L2 learning” (p. 3). Moreover, there is a general consensus that the attention to input establishes the internalization of language (e.g., Long, 1991; Schmidt, 1990, 1993, 1994, 1995, 2001; Sharwood Smith, 1988; Slobin, 1985; Tomlin & Villa, 1994; VanPatten, 1990, 1994; Rosa & O'Neill, 1999). As a crucial construct for SLA, attention therefore seems to be of paramount importance just as input is (Simard & Wong, 2001; Wong).

As the essential role of input has become widely recognized in the field, and a rich corpus has similarly explored the input-to-intake phenomenon, many insights into language pedagogy have been gained as well. Hence, VanPatten (1996) urges that longstanding teaching practices, especially in grammar, be reevaluated owing to the developments in this aspect of SLA. He debates “whether traditional grammar instruction is consonant with the idea that input is the basic building block for the construction of a mental representation of the L2 grammar” (p. 5). Cadierno (1995) likewise directs the issue away from “whether grammar should be taught …[towards] how it should be taught” (p. 190, emphasis original). On the same vein, Ellis (1998a) poses one of the key questions that language teaching specialists have pursued for more than 30 years now: “How can we teach grammar in a way that is compatible with how learners acquire grammar” (p. 1). More engagingly, Ellis (1997) aptly articulates what language teachers have long wanted to discover: “[W]hat kind of grammar instruction works best[?]”(p. 77).

In retrospect, the standard grammar instruction by tradition was captured in essence by what came to be known as the PPP model (Figure 1). In this framework, lessons would usually begin with presentation, which took the form of a grammar explanation, followed by practice through exercises, until learners produce accurate answers. Production in turn was meant for making the learners use the second language fluently. In other words, this paradigm works on the premise that “accuracy precedes fluency” (Thornbury, 1999, p. 28).

Figure 1. PPP Model.

Thornbury, on the contrary, asserts:

All language learners go through a stage of making mistakes; meanwhile, they may be perfectly capable of conveying their intended meaning fluently…. Delaying communication until accuracy is achieved may be counterproductive. Rather than as preparation for communication, it seems that it is by means of communication that learner's language system establishes itself and develops” (p. 129).

On top of this incompatibility, VanPatten (1996) points out that most traditional approaches to grammar instruction are output-based, in which case learners are made to produce correct forms right at the outset of a grammar lesson (Figure 2). He argues that these practices are incongruent with the recent generally accepted research on L2 input. The output practice or the explanation about grammar or the textbook is not the kind of language data ready for intake derivation of the developing system. Moreover, Ellis (1998a) concurs with this, noting that “current theories of SLA see production as the result of acquisition rather than the cause. It follows that grammar be taught more effectively through input rather than through manipulating output” (p. 2).

Figure 2. Traditional Practice in Grammar Instruction (taken from Lee & VanPatten, 1995, p. 95).

Although a number of SLA researchers and applied linguists (e.g., Dulay & Burt, 1973; Krashen, 1982; Prabhu, 1987, as cited in Ellis, 1997) challenge form-focused instruction and advocate untutored communication-rich settings instead, Ellis (1998b) is positive that efforts at teaching grammar can succeed, provided that certain psycholinguistic aspects are taken into account. Offering a framework of what he calls methodological options that relate to the psycholinguistic processes and constraints relevant to L2 acquisition, he suggests that an option, therefore, be “ideally be psycholinguistically motivated” (p. 77).

Lee and VanPatten (1995), on the same note, maintain this pedagogical outlook by offering “a way to incorporate explicit grammar instruction into classes without sacrificing either communication or learner-centered activities” (p. 94). Consistent with both SLA theories and communicative language teaching (CLT) principles, Processing Instruction (PI) is a type of grammar instruction designed to affect ways in which learners attend to input data, as first described in VanPatten (1990) and has since been extensively discussed in numerous publications (e.g., Cadierno, 1992, 1995; Cheng 1995, as cited in Van Patten 2002; Lee & VanPatten, 1995; VanPatten, 1996, 2002, 2004; VanPatten & Cadierno, 1993; VanPatten & Oikkenon, 1996; VanPatten & Sanz, 1995).

Van Patten then suggests that language specialists and/or teachers go about their grammar lessons at the level of processing input (Figure 3). The use of PI aims to “get learners to attend to grammatical data in the input and to process it” (Lee & VanPatten, p. 102). This is predicated on the goal of “alter[ing] the processing strategies that learners take to the task of comprehension and to encourage them to make better form-meaning connections than they would if left to their own devices” (VanPatten, 1996, p. 60). Basically, among its other characteristics, it attempts to “push learners away from the nonoptimal strategies during input processing” (VanPatten, 2002, p. 764) by utilizing a particular type of input known as structured input. Further, being an input processing-based instruction, it does not require the learners to produce the targeted grammatical form during the PI segment of grammar teaching. Its primary objective is to make the learners attend to form while they engage in making out sentences. Conversely, the production of the form in question or the role of output, on the other hand, is put aside for the purposes of developing fluency and accuracy (Lee & VanPatten; VanPatten, 2002).

Figure 3. Processing-Oriented Grammar Instruction (taken from Lee & VanPatten, 1995, p. 99).

The present research agenda has, therefore, been motivated by this relatively fresh pedagogical approach that would hopefully address the contradiction between the recent theories in L2 learning and the actual language classroom practices, which creates a wider gap between the language acquisition-language teaching connection. It may be said that this theory-to-practice disparity is evident far and wide. In the United States, for instance, language teaching methods that are hardly communicative, or akin to the old production practice, still thrive (VanPatten, 1996). Schulz (1991, as cited in Magnan, 1991) shares the same observation by pointing out that “our textbooks lag far behind our theoretical desires” (p. 324).

It goes without saying that English language teaching in the Philippine classrooms closely parallels that in the U.S. As evidenced by local research and the vast majority of textbooks, let alone the personal observations of the researchers of this paper, the staples of grammar teaching have been heavily flavored by the PPP model while CLT seems overall a nascent menu. Incidentally, a good many Filipino learners of English as a second language (ESL), even those at the tertiary level, manifest persistent grammatical errors in that language; in this particular case, the English present perfect is a perennial common problem. It may be hypothesized that the perceived low level of their linguistic competence can be ascribed generally to the prevailing pedagogical environment.

On a positive note, an “informed approach” (Brown, 1994, p. 73) to language learning can redirect the “changing winds and shifting sands” (Marckwardt, 1972, as cited in Brown, 1994, p. 52) of the trends in language teaching over the years. The psycholinguistic underpinnings of processing-oriented grammar instruction, advocated by VanPatten and his colleagues, have beckoned with optimism to bridge the gap between theory and practice. VanPatten (1996) strongly believes that the input-based grammar instruction, such as Processing Instruction (PI), is superior to its traditional output-based counterpart, in the light of more modern theory and research. It is worth mentioning at this point that the beneficial effects of PI is due to one of its basic features or key components – the structured input activities (e.g., Benati, 2001; Sanz & Morgan-Short, 2001, as cited in Vanpatten, 2002; VanPatten & Cadierno, 1993; VanPatten & Oikkenon, 1996;). Therefore, this approach to grammar teaching can be exploited in the ESL classroom in order to find out if and how it works among Filipino L2 learners.

Taking the pedagogical concerns into account, PI is deemed to be a viable methodological option – one understood to be “psycholinguistically motivated focus on form [intervention] that is an adjunct to communicative language teaching and/or comprehension-based approaches” (Sanz & VanPatten, p. 1). Hopefully, if proven its effectiveness, this alternative has yet to be a more effective tool from which the ESL teachers and students alike will stand to benefit.

Therefore, in light of these developments in the teaching of grammar, this paper aimed at shedding light on the following questions

• Do TI (traditional instruction) and PI have any effect(s) on the interpretation and production of English present perfect of Filipino college students?
• Does PI bring about better performance of the students in learning the said grammatical item, as compared with that of the more traditional, outputbased grammar instruction?

For more than a decade now, a growing corpus of SLA research informs on the effects of certain types of formfocused instruction in second language (L2) (e.g., Celce- Murcia, 1991; Celce-Murcia, Dornyei, & Thurrell, 1997; Doughty & Williams, 1998; R. Ellis, 1993, 1998: N. Ellis & Laporte, 1997; Fotos, 1994; Fotos & R. Ellis, 1991; Kondo- Brown, 2001; Long & Crookes, 1992; Loschky & Bley- Vroman, 1993; Nunan, 1989; Savignon, 1991; Spada, 1997; Terrell, 1991; VanPatten, 1996, 2002, 2004; Williams, 1995). For one, recent studies aim to identify the way(s) of attending to form in meaning-based or task-based communicative language teaching (CLT). For another, further researches attempt to determine the relative effects and/or superiority of one type of practice over other types (Kondo-Brown, 2001).

In particular, the seminal work of VanPatten and Cadierno (1993a) spawned the line of research into Processing Instruction (PI) and several replication studies. It set out to investigate the processing of input by second language (L2) learners through explicit instruction. The researchers had the learners modify their strategy in making formmeaning connection during exposure to input strings in a tutored setting. Clearly, this approach deviates from the traditional explicit instruction, in which the learners are made to generate language items right at the outset of the lesson procedure. The study then attempted to shed light on the following research questions

• Does altering the way in which learners process input have an effect on their developmental systems?
• If there is an effect, is it limited solely to processing more input or does instruction in input processing [IP] also have an effect on output?
• If there is an effect, is it the same effect that traditional instruction [TI] has (assuming an effect for the latter)? (VanPatten 1996, p. 88).

VanPatten and Cadierno conducted an experiment to test the Input Processing (IP) Theory using the word order and object pronouns in Spanish.

In Spanish, object pronouns precede finite verbs, and subjects may be optionally deleted or may appear postverbally. Previous research had demonstrated that learners of Spanish misinterpreted object-verb-subject (OVS) and object-verb (OV) structures as SVO and subject-verb (SV) structures, respectively (VanPatten, 2002, p. 769).

It involved three groups of learners, namely, a Processing Instruction (PI) group, a traditional instruction (TI) group for purposes of comparison.

It was found out that the PI group “made significant gains” (VanPatten, 2002, p. 771) on the two assessment tools of the study: a sentence-level interpretation test and a sentence-level production test. On the other hand, the TI group did so only on the production test, whereas the control group did not make any gain at all.

With reference to the research questions, the researchers inferred

First, altering the way learners process input can alter their developing systems. The processing group showed evidence of this on both interpretation and production tests. Second, the effects of PI are not limited to processing but also showed up on production measures. Finally, the effects of PI are different from those of TI. With PI, learners get two for one. By being pushed to process form and meaning simultaneously, they not only became able to process better but could also access their newfound knowledge to produce a structure they never produced during the treatment phase. The traditional group made gains only on production and did not make gains in the ability to correctly process form and meaning in the input. The researchers took these latter results to mean that the TI group learned to do a task, whereas the PI group experienced a change in their underlying knowledge that allowed them to perform on different kinds of tasks. They concluded that instruction was directed at intervening in the learners' processing strategies should have a significant impact on the learners' developing system (VanPatten, 2002, pp. 771-772).

Meanwhile, the foregoing initial study, with its understandable limitations, prompted the self-same researchers and others alike to widen and explore further the research agenda. They then posed the following series of questions :

• Are the effects generalizable to other structures?
• Are the effects of PI due to different explicit information?
• Are the effects of PI observable with different assessment tasks?
• Are the effects of PI different from those of other types of instruction? (VanPatten, 1996, pp. 88-121; VanPatten, 2002, pp. 772-789).

Cadierno (1995) made a replication study, this time, focusing on the Spanish preterit (past) tense, deemed to be a more challenging structure for learners considering its complexity in form. The results of the study were the same as that of VanPatten and Cadierno (1993a). It, therefore, reaffirmed the relatively more beneficial effects of PI than TI's.

Using other grammar items in Spanish, Cheng (1995, as cited in VanPatten, 2002) replicated VanPatten and Cadierno's (1993a) study, focusing on ser and estar, the two major copular verbs in Spanish. Moreover, she included a more complex test by virtue of having the subjects of the study write a composition after taking a look at certain pictures (besides those interpretation and production assessment tools that VanPatten and Cadiero had administered). Cheng's results paralleled those of VanPatten and Cadierno. It showed that only the PI group performed significantly well on the interpretation test, while both the PI and TI groups made progress on the production test, and “were not significantly different from each other” (VanPatten, p. 773). All the PI, TI and control groups, however, “improved significantly” (p. 773), but the PI and TI groups were significantly better than the control group as shown in their respective posttest scores. It should also be noted that “[t]here was no difference between the processing and traditional groups on the composition tasks” (p. 773).

In addition, Farley (2001a, as cited in VanPatten, 2002) confirmed the beneficial effects of PI on the subjunctive in Spanish with noun clauses. In his research, the participants under the PI treatment showed significant improvement in their learning of the target structure in both form and function. Furthermore, the positive results were noted not only in the interpretation but also in the production measures.

In another study, Buck (2000, as cited in VanPatten, 2002) attempted to find out which was more effective between PI and TI in the acquisition of the present continuous in English by native speakers of Spanish. Her study yielded indication of “greater gains for the PI group that were maintained over time on the interpretation test; initial gains made by the TI group were not maintained” (p. 773), however. There were “similar gains,” (p. 773) though, in the two groups on the production test and were evident over the same extended period.

The superiority of PI over TI, at least, was again made manifest by VanPatten and Wong's (2004) study. With the French causative (faire causative) as the target structure, their results mirrored those of VanPatten and Cadierno (1993a). The PI group improved significantly better than the TI and the control groups in the interpretation test; the same experimental group did as well as the TI group in the production test.

In the same vein, the relative effects of PI and TI on the acquisition of verbal morphology in the future tense in Italian became another replication inquiry by Benati (2001). This time a pretest, an immediate post-test, and a three-week delayed test formed part of the research procedure. Also, it is worth mentioning that his was not like VanPatten and Cadierno's (1993a), in that “mechanical or purely form-activities” (VanPatten 2002, p. 775) were hardly ever utilized. The results “were similar to, but not the same as, those of the original VanPatten and Cadierno's (1993a) study” (VanPatten, 2002, p. 774). Greater gains were made by the PI group on the interpretation task while both the PI and the TI groups improved on the two production tasks without difference between the two.

In summary, the foregoing replication studies suggest that the Input Processing (IP) Theory, in particular, the learners' processing strategies for form-focused instruction, and the effects of PI are generalizable to structures and languages other than those examined in VanPatten and Cadierno's original study (VanPatten, 2002), to wit.

There is evidence for the superiority of PI over TI for pronouns and word order in Spanish, for complex verbal morphology in Spanish (the preterit) and Italian (future tense), for lexical-aspectual choice (copular verbs in Spanish, for agent-dative relations and word order in French, for mood selection in subordinate clauses in Spanish, and for the present continuous versus the progressive in English (p. 775).

However, another limitation of VanPatten and Cadierno's (1993a) initial study compelled researchers to do a followup, in order to rule out the significant contributory factor of explicit information to the effects of PI, if any. It can be recalled that the procedure PI involves an explanation on the target grammatical item, information about the processing strategies learners apply, which may not work to their benefit, and structured input activities. On the other hand, TI follows the traditional grammatical explanation of the structure without giving information on any processing strategies. Hence, an attempt to “tease out explicit explanation as a variable” (VanPatten, 2002, p. 786).

Consequently, VanPatten and Oikennon (1996) furthered the line of inquiry by testing their hypothesis that the explanation-giving feature-segment of PI did not contribute anything significant to the positive outcomes of the participants' language learning described previously. The results, interestingly, led the researchers to point to “the particular nature of the structured input activities, and how these pushed [the PI group and the structured-input-only group] to make form-meaning connections because the input sentences have been manipulated in a particular way” (VanPatten, 2002, p 786), thus concluding that the effects of PI were not attributable to the explicit information given to the learners.

The strength of PI's structured input activities was later supported by another study by Sanz and Morgan-Short (2001, as cited in VanPatten, 2002), utilizing computerassisted language learning (CALL). In this case, however, the investigation sought to find out whether the component of explicit feedback – CALL's selling point – is essential or advantageous to learners. Four groups were compared employing the variables [+/- explicit explanation] and [+/- feedback]. The results showed that all the groups performed well on the interpretation task and the production task with no significant difference between and among them. They concluded that “neither explicit information nor explicit feedback seemed to be crucial for a change in performance; practice in decoding structured input… alone seems to be sufficient” (p. 787).

Similarly, Benati (2004) validated the crucial or more important role for the structured input activities of PI than the one for explicit information. In his study using the future tense in Italian, the PI group and the structured-input-only group showed significant improvement much more than the explicit-information-only group on the pretest and posttest and on the interpretation and production tasks.

In an attempt to measure the extent of the relative effects of PI to the discourse level (since VanPatten and Cadierno (1993a) used only sentence-level tests while Cheng utilized a composition task), VanPatten and Sanz (1995) employed three kinds of output tests: a sentence-level test, a questionand- answer (based on pictures) test, and a video narration test in both oral and written modes. The researchers intended to find out the outcome of the video narration task more importantly “because of the complex cognitive activity it represents … [in which] the participants must provide all vocabulary, all syntax, and all grammatical features on their own without prompts” (VanPatten, 2002, p. 788). The study yielded better gains by the PI group especially in the written mode than that of the control group.

Along those lines, Sanz and Morgan-Short (2001, as cited in VanPatten, 2002) computer-assisted study utilized the same assessment tasks as those of VanPatten and Sanz (1995). The participants similarly demonstrated a significant improvement on every task; needless to say, they did just as well on the video retelling.

VanPatten (2002), therefore, posited that not only do output kinds of tests on the sentence level determine the relative effects of PI, but those on the level of discourse apply as well to that purpose. Moreover, written tests tend to yield better performance than do oral tests.

To determine whether the effects of PI are different from those of other types of instruction, Farley (2001a, as cited in VanPatten, 2002) pursued this inquiry furthermore. He compared PI with “meaning-based output” instruction (MOI) with PI. MOI follows the principles of structured output activities by having no mechanical drills involved as well, as discussed in Lee and VanPatten (1995) and in VanPatten and Cadierno (1993a). It was found out the PI group performed just as well as the MOI group. Therefore, neither type of instruction was superior to the other. Although it should be noted that in a follow-up study by Farley (2001b, as cited in VanPatten, 2002), its results did not match the same findings of the previous research. Retaining the design, procedure, and target structure in the Farley (2001a) study, only the PI group performed consistently well on a delayed task; the MOI did not so. This consequently proved the superiority of PI to TI as previously validated by Buck (2000, as cited in VanPatten, 2002).

More recently, Morgan-Short and Bowden (2006) studied the effects of meaningful input- and output-based practice in language teaching. Forty-five (45) Spanish students received different instructions under processing instruction, meaningful output-based instruction and control groups. The same input was given to the two experimental groups namely processing instruction and meaningful output-based instruction. The difference rested on the meaningful practice: input based for the processing instruction and output-based for the other experimental group. Results that were subjected to repeated-measures analyses of variance showed significant gains on immediate and delayed interpretation and production tests for the experimental groups. Furthermore, results indicated that both experimental groups outperformed the control group for all the interpretation tasks. For the production tasks, only the meaningful output-based group outperformed the control group. The results of their study suggest that both input- and output-based instructions can help in the learners' linguistic development.

In conclusion, the aforementioned related studies speak of the superiority of PI over TI and its corollary – the relative effectiveness in grammar instruction. It can be inferred that effects of PI are generalizable to other structures and to those in other languages for that matter; that these effects are due mainly to the structured input activities as one key component of PI; that the same exhibit in different assessment tasks, from sentence to discourse levels, and both in the oral mode and especially so in the written one; and, that the significant gains of PI experimental group(s) manifest overtime. Therefore, the findings of VanPatten and Cadierno (1993a), whose original study had served as “the impetus that launched [this] research agenda” (p. 769), turn full circle as the replication studies have validated the theory of IP substantially.

It must be pointed out, nevertheless, that there are other empirical studies yielding results different from those discussed earlier (e.g., Allen, 2000; Collentine, 1998; DeKeyser & Sokalsi, 1996; Nagata, 1998; Salaberry, 1997). The results apparently have fallen short of establishing the relative effectiveness and/or any superiority of PI and/or structured input activities over other types of grammar teaching practices (Kondo-Brown, 2001). VanPatten (2002) addresses this anomalous discrepancy by claiming that these studies were not genuine replications of his original, in that they concern some methodological differences. Therefore, such issues are deemed beyond the scope of this chapter. See DeKeyser, Salaberry, Kondo-Brown, Salaberry (1998), Sanz and VanPatten (1998), and VanPatten for discussions of these issues at length. Kondo- Brown also outlines the seminal studies that examined the relative effects of input-oriented versus output oriented grammar instruction.

In the local school setting, Erfe (2006) aimed to develop structured input activities in English grammar instruction for college freshmen in one public university in Manila. Initially, he specifically attempted to identify common grammatical errors in English of first year college students, for which he prepare structured input activities, following the guidelines in Lee & VanPatten (1995) and VanPatten (1996, 2002). Then, he undertook a formative evaluation of the activities through a content review by colleagues in the field, and a field try-out of the activities on the target users. Finally, he revised the field-tested activities guided by the results of the evaluation.

Erfe (2006) found that the guidelines for developing structured input activities and the major types of these activities illustrated in Lee and VanPatten (1995) and VanPatten (1996, 2002) were useful for preparing instructional materials in English that show potential to address the need for psycholinguistically motivated approach to second language (L2) grammar instruction. He also inferred that these structured input activities could be used to help Filipino learners of ESL to focus on formmeaning connection in L2 grammar instruction. Thus, the present study is a further experimental inquiry into the relative effect(s) of PI, in which structured input activities is a major essential component. This academic endeavor was launched in tandem with his able colleague who is also into university teaching.

Eighty (80) sophomore college students served as the subjects of this experimental study. They were enrolled in a second-semester oral communication course in English (with integrated grammar lessons) in a coeducational private university in Manila. Their ages ranged from 17 to 19 years, and whose official admission to the university, upon placement screening, and successive promotion to second-year level bore out their average–above average mental ability, English language proficiency, and scholastic achievement. With Filipino (the Philippine national language) as their L1, they had formally learned English as a second language since preschool, although the majority would still manifest some L2 grammar errors. These students constituted two classes that were randomly assigned to two (2) groups for the present study: the experimental group (n=40), receiving Processing Instruction (PI), and the control group (n=40), receiving traditional, output-based instruction (TI).

The present perfect aspect in English was chosen as the morpho-syntactic item for instruction given to the subjects of the study. This aspect is particularly problematic among Filipino learners of English, even at the tertiary/university level. The subjects were assumed to have not assimilated such L2 grammar structure in their developing system, as evidenced in their teachers' formative and summative assessments, common observation of the language output of the students, and the results of a standardized multiple-choice grammar-structure test administered to the students prior to the instructional treatment. The grammar test likewise revealed that the present perfect ranked first among the most number of errors made by the students. Incidentally, the same structure was part of the lessons in the course they were attending.

Of the usual four (4) similar functions of the present perfect, the experiment focused only on one (1) use of the present perfect, that is, to express a situation that began in the past at a prior point in time and continues into the present, as in the following sentence:

I have taught English since 1998.

Besides time constraints for accommodating the conduct of the study involving the students during regular class, this was in consonance with one of VanPatten's guidelines, that is, “present[ing] one thing at a time” in a given lesson. The instructional treatment took an overall duration of one and a half hours (1 ½), scheduled for three separate sessions in a week.

The pretest-posttest design was employed in the present study. Both the experimental group and the control group undertook the same pretest and post-test, except that the experimental group was taught the present perfect utilizing Processing Instruction. The control group, on the other hand, received the same lesson following the usual teaching procedure: presentation of the target structure via lecture, practice using the grammatical form and function through worksheets, and production of written output in expository form applying the same structure. The pretesting, which likewise purposively determined the equality of baseline knowledge of both groups, was done one (1) week before the instructional treatment. The posttest (immediate) was administered to the same groups two (2) days after the instruction phase, and the self-same test (delayed) was again given them two (2) weeks thereafter. Results of these tests were analyzed using paired-samples T test for determining the significant differences of the means of the pre- and post-tests, and independentsamples T-test for comparing the TI and PI group performances.

For the experimental group, their instruction was designed and implemented based on the three (3) basic features or key components of PI (VanPatten, 1996; VanPatten, 2002). First, the teacher conducted in class a lecture on the structure of the present perfect (i.e., has/have + past participle form of the verb) and its corresponding meaning (i.e., a situation that began in the past at a prior point in time and continues into the present).

Next, the teacher made the students aware of certain strategies that they had to be cautious about when presented with the target structure. To optimize their learning of the form and meaning of the present perfect, they were reminded not to rely on time markers expressed usually in the adverbials of sentences. Such strategy would tend to defeat the purpose of leading the students to make “form-meaning” connections. Instead they should pay attention to the verb form that carries the aspectual meaning of a given sentence. For example, in the sentence The University has produced luminaries in the arts and sciences for four centuries now, they had to take their cue from the present perfect structure itself (i.e., “has produced”) for grasping that the action was ongoing or has relevance to the present time, and not from the time adverbial “for four centuries now.” The same would be true for making out the sentence The College of Architecture was awarded Center of Excellence in 2001. The students should focus on the verb phrase, not on the adverbial of the sentence, to interpret the “pastness” of the action.

Lastly, the students engaged in structured input activities, which had previously been developed and validated by the researchers. These activities were designed to prompt the students to focus on the targeted grammatical form of the lesson, instead of other lexical cues, in the language input. For the purposes of this experiment, this component of PI was composed of five (5) structured input activities for the particular grammar lesson. The activities also used a combination of referentially oriented activities and affectively oriented activities. Referential structured input activities were those of objective-type, with correct or incorrect answers, that necessitated paying close attention to the form for meaning.

The students were likewise made to respond to affective structured input activities, creating a meaningful learnercentered communicative language teaching (CLT) classroom. These were activities in which they expressed their “opinions, beliefs, feelings, and personal circumstances,” which provided processing information about themselves and the real world (VanPatten, 1996, p. 64). Unlike referentially oriented activities, affectively oriented activities had no wrong or right answers. Moreover, the activities operated from sentence to longer stretches of discourse as they progressed.

On the other hand, the control group went about the lesson on the present perfect aspect, with the typical instructional procedure of their teacher, at the same time and duration as that of the experimental group. The targeted grammar structure, along with its meaning, was presented to the students. The students were then asked to answer relevant written tasks in their workbook, such as sentence completion and generation of original sentences using the present perfect. The instruction culminated in an expository writing that required the use of the same structure.

A pre-test and a post-test, which were identical in format and number, were prepared for the study. Each consisted of interpretation tasks in which the students had to underline the present perfect form in a given expository text to test their receptive knowledge of the target structure. For another part, a grammaticality judgment component of the test was given, for them to determine whether the sentences were morphologically and syntactically correct; if not, they had to revise the verb forms to correct the sentences. Moreover, the students were also made to engage in production tasks, such as supplying the correct verb forms of sentences, generating original sentences using the present perfect, and writing a short-essay about their personal experience at the University, in which they should be able to employ the present perfect aspect in at least three (3) sentences in their composition. These tasks aimed at assessing their ability in the actual use of the structure as well. It should be noted that the test items varied from discrete-point to discourse-level types, which also mirrored the structures input activities attempt at “moving from sentence to connected discourse.” The simple past tense was also incorporated in several items of the test to serve as distractors.

Every correct response to each item in the interpretation tasks was accorded one (1) point. For the production tasks, every correct use of the present perfect in the written discourse earned one point as well. There were 10 points for the interpretation tasks and 20 points for the production tasks.

Paired-Samples T Test was employed in the interpretation of the TI and PI groups' scores to determine answers to the first research question: Do PI and traditional, output-based instruction have any effect(s) on the interpretation and production of English present perfect of Filipino college students?

Table 1 presents the differences of scores of four paired tests for the TI group: pre-interpretation and post interpretation; pre-interpretation and delayed interpretation; pre-production and post production; and, pre-production and delayed production. Results revealed that at 5% level of significance, there are considerable differences in the mean performance of the respondents from pre-production to post production at .000 p value and from pre-production to delayed production tests at 0.000 p value. These figures indicate that the TI group's post production and delayed production test scores are significantly higher than their pre-production scores. In other words, the TI intervention resulted in improved positive scores on the students' post production and delayed production tests.

Table 1. Paired-Samples ‘T’ Test Results for the TI Group.

Results of the paired-samples ‘t’ test results for the PI group (Table 2) revealed statistically significant scores of the group's delayed interpretation (p value= .005), post production (p value= .000) and delayed production (p value= .000) tests. The group posted a mean score of 8.3500 delayed interpretation from 7.5750 preinterpretation score, a mean difference of -.77500. Moreover, the group improved its post production score of 17.5000 and delayed production mean score of 17.5250 from 13.1750 pre-production test.

Table 2. Paired-Samples ‘T’ Test Results for the PI Group.

The results for the first research query then are as follows: (a) on the interpretation task, only the PI group made significant gains in the delayed interpretation test, and (b) on the production task, both groups yielded statistically favorable gains on the post production and delayed production tests.

The next research question set out was, “Does PI bring about better performance of the students in learning the said grammatical item, as compared with that of the more traditional, output-based grammar instruction?” To answer this question, independent samples ‘t’ test was utilized to compare the performance of the two groups.

Results of the independent t test for the TI and PI groups in the pre-interpretation and pre-production test yielded no significant differences at p value = .940 and p value = .668, respectively (Table 3). These figures indicate that there is no difference between the groups before instruction and thus, establish homogeneity of the two groups in this study.

Table 3. Independent ‘T’ Test Results for the TI and PI Groups.

Moreover, the ‘t’ test for equality of means show the following statistically significant findings: p value = .037 for the delayed interpretation, p value = .000 for the post production and p value= .000 for the delayed production. These values show statistically relevant differences in the scores of the TI and PI groups; that is, the PI group posted higher scores in those three tests.

The second research question which delves into the comparability of the performance of the two groups yield the following answers: on interpretation task, PI outperformed the control group in the delayed interpretation test (p = .037); whereas, on the production task, the PI group posted more superior results over the TI both in the post production (p = 0.000) and delayed production tests (p = 0.000).

In light of these results, it appears that PI has significant effects on students' learning of the English present perfect aspect. In particular, input processing, being the core of PI, seems to reinforce the linguistic knowledge of the students via retention of that input in their developing system. Similarly, the results of this study validated further the initial findings obtained by VanPatten & Cadierno (1993a) and Cadierno (1995) where the PI group outperformed its traditional counterpart on interpretation, and both groups made significant gains in the production task.

The present study explored the effects of traditional, outputbased and processing instruction methods in teaching the English present perfect aspect. Based on the results presented above, both the TI and PI have positive effects on improving the production performances of the groups as shown in the significant gains of the learners post production (p value = .000) and delayed production (p value = .000). Moreover, only the PI group yielded better performance on their delayed interpretation task at p value = .005.

Furthermore, it appears that PI yielded more significant gains on the improvement of the present perfect aspect of the students. This is revealed by the statistically relevant differences in scores showing the superiority of PI over TI in delayed interpretation (p = .037), post production (p= 0.000) and delayed production (p = .000) tasks.

The significant gains achieved by the students of the PI group in outperforming their TI counterpart may then be attributed to the features of the processing instruction. It might be assumed that the second component namely, information of learning strategies that might work against their benefit has raised their awareness in assimilating input during instruction. Hence, they became extra-careful when they worked on the structured-input activities, which helped them further to focus on the targeted linguistic item for acquisition.

Indeed, SLA research has significantly contributed to an understanding of the intricacies of language learning, as the interesting probe into the role of explicit instruction in the same has also come a long way. According to Braidi (1999), “[t]he results of research can add to the knowledge base from which teachers make informed decisions: to formulate, or reformulate, their notions about language, language learning, and language teaching” (p184). As Lightbown, Spada and White (1993) likewise observe, it has since advanced the knowledge of how input bears on acquisition. Therefore, recent studies centering on the input theory should offer useful implications for language teaching, in order to shed light on the two most fundamental research questions in the field

• Through what processes do learners learn an L2?
• How can teachers best enable and support those processes? (Macaro, 2003, p. 21)

Clearly VanPatten and his associates have responded actively to such relevant issues and blazed a new trail of pedagogical inquiry, which the researchers of the present study sought to follow. VanPatten stresses, however, that PI need not replace the whole of the language program; the other language arts (i.e., listening, speaking, reading, writing, etc.) should still be explored to operate in harmony, within the parameters of communicative language teaching.