Abstract

This study has investigated the relationship between the age of Iranian EFL learners and the strategy they seek in their interpretation of sentences. A task was devised to solicit either prosodically or syntactically motivated responses from two groups of participants aged 12-13, and 17 and over, respectively. Comparison of the different age groups' performance on the task revealed that younger learners were more inclined to follow prosodic cues and older learners were more inclined to follow syntactic ones.

Key Words: Age, Processing, Prosody, Syntax, Cues.

1. Introduction

1.1. Situation

Teacher: "Suppose your mother has poured you a glass of milk. After a few minutes, she calls out wanting to know if you have finished your milk or not.
Would you say that the glass is half empty or full?"
Student: "It depends on whether you like milk or not?"
Teacher: "In what way?"
Student: "Well, if you like milk you'd say it's half empty; so she'd pour you more milk; if not, you'd say it's half full; so she wouldn't".

(An actual word-for-word reproduction of an oral interaction between the author and a twelve-year old EFL learner, as unfolded at one of the 'Iran Language Institutes' lower-intermediate classes in the Summer of 2005).

There have been innumerable instances of cases of this nature - the baffling spontaneity and analytical prowess of the younger learner's inter-personal communicative L2 skills - which have prompted the formulation of the ensuing empirical study. Casual observation of the adult learner's somewhat disappointing communicative performance in spite of his far superior edge in academic achievement tests has decidedly provided a rather profound impetus for an empirical study. Notwithstanding the subjectivity from which the above-mentioned observations suffer, they nevertheless mirror the long-established dogma that has commonly been associated with the merits of early age exposure to L2. It has been observed by Stern (1983, p.361) that many notable educators like Erasmus, Montaigne, and Locke have over the centuries argued in favor of "an early age start in second language learning".

The explanatory hypotheses offered in this connection - the merits of early age exposure to L2 - have by and large relied on such diverse grounds as: " biological factors, affective factors, motivation, time allotment, cerebral dominance (hemisphericity), and learning conditions" (Larsen Freeman, 1997, p.175), not to mention language input.

Empirical studies dealing with age-related variations in L2 learning ability, too, have for the most part left much to be desired in unanimity of consensus. Some researchers like Asher & Garcia (1969) and Seliger, Krashen & Ladefoged (1975) have found that in the long run younger learners are more likely to attain near-native L2 proficiency in the more natural (ESL) situations. Others, such as Ekstrand (1976, 1978), and Fathman (1975), have claimed that when time is held constant older learners tend to syntactically outperform the younger ones in both natural and planned settings.

These contradictory findings have led to the assertion that there might be some processing dimensions at play which could help account for the rate and degree of L2 acquisition at different ages. In this light, it is the aim of this study to seek empirical evidence for age-related processing preferences of L2 learners.

1.2. Justifications
The theoretical justifications offered for the causes of the said antithetical findings have, on the one hand, revolved around the rationalistic notion of the existence of a 'critical period' in the learner's life which affects his language learning ability. Proponents of this theory attribute the younger learner's eventual success in L2 proficiency to the plasticity of his brain (Krashen, 1982). Theorists of other persuasions, the cognitivists, however, maintain that older learners are more cognitively equipped to handle syntax (Dulay, Burt, & Krashen, 1982).

From an empirical point of view, however, it seems reasonable to suggest that the younger learner's more respectable overall performance in the long run, if not a result of being more fortunately circumstanced syntactically, must be, all things being equal, attributed to some hitherto unknown processing advantage.

Amongst the many processing dimensions needing empirical research, the extent of the learner's attentiveness to prosodic cues, as opposed to the syntactic ones, tends to provide the most reliable and valid of all platforms for the task at hand. Such a platform is deemed reliable on the account of not being as exposure dependent as the post-processing dimensions of success or product like, for instance, pronunciation which is after all enhanced with experience. As for validity considerations, the present study's platform seems to be remarkably compatible since it juxtaposes and contrasts the mentioned cues as they compete, in the course of the same sentence, for the learner's attention.

1.3. Objectives
The hypothetical framework on which this study is based has been derived from studies conducted on L1 processing. The results obtained by Fernald & Mazzie (1991) and Read & Schreiber (1982) confirm the existence of an age-related processing dimension in L1 acquisition. More specifically, it was found that younger speakers are more inclined to attend to prosodic cues of structure, while older speakers relied more on syntactic cues.

In the realm of L2, only one study - Harley (1995) - has tried to shed some light on age-related processing variations of foreign language learners. Harley's (1995) findings, however, due to some unforeseen shortcomings in sample selection, failed to uphold the hypothesis that younger FL learners are more influenced by prosody and older ones by syntax. Guided by these findings, this study seeks to ascertain age-related variations in L2 processing via the implementation of a task which exacts a measure on the learner's attentiveness to prosodic versus syntactic cues of sentence structure.

1.4. Significance
Should the outcome of this study verify the hypothesis on which it was postulated - namely, that younger learners are more attentive to prosody than syntax and the reverse order to be true for older learners - it would have far-reaching theoretical and pedagogical implications. First and foremost, given the cause and effect relationship generally held to exist between speech perception and production; pinpointing such age-related differences in L2 processing not only accounts for the younger learner's alleged superiority in listening comprehension but also for his eventual near-native L2 production, as noted by Burstall et al. (1974) and Selieger etal (1975), respectively. Furthermore, exposing the potential strengths and weaknesses of different age categories can aid educators in developing appropriate syllabi and teaching techniques.

2.0. Review of Literature
2.1. Theoretical Perspectives
The theories addressing language acquisition in general and the age factor in particular are dispersed along a rationalist-empiricist continuum. At one end of this continuum lies the rationalistic view that language is to a large extent specified biologically, or innately; and on the opposing end, in stark contrast, the role of experience is emphasized through the empiricistic vantage point.

Under the predication of the rationalistic perspectives, according to Harley (1986), language is considered a by-product of neurobiological factors; while, the empiricists maintain that environmental factors are more instrumental in bringing language to bear. Given the long history of heated debates between proponents of the said persuasions - with no clear edge gained by either side - a good number of theorists have opted for more conciliatory, mid-continuum platforms in establishing their theoretical constructs. Thus, the following elaboration on theoretical perspectives begins with those on the rationalistic end of the continuum and moves toward the more empirically inclined ones as it unfolds.

2.1.1. Brain Plasticity Hypothesis
In so far as neurological evidence is concerned, the forum that has most effectively taken advantage of it is the concept of the 'Lateralization of the brain'. Lyons (1981, p. 249) views this phenomenon as the hemispherical specialization of the brain to perform specific tasks. He further maintains that "process of lateralization is maturational, in the sense that it is genetically preprogrammed but takes time to complete".

The hypothesis put forth by the proponents of lateralization, supported by clinical studies of aphasia, relates the ease with which children acquire language with the so-called "plasticity" of their brains. Penfield & Roberts (1959), as cited by Harley (1986), maintain that the child's brain is more efficient in processing language due to its far greater flexibility, or plasticity, as compared to that of an adult's. It is argued that the child exhibits a sort of natural predisposition to acquire language before the age of nine to twelve. Moreover, according to Harley (1986), Penfield & Roberts claimed that due to a more plastic brain the child has an advantage over his adult counterpart in acquiring the early sets or units of language, while the adult is probably more advantaged in the area of vocabulary development.

2.1.2. A Biologically Based Critical Period
There is ample evidence to believe that language acquisition is more a function of the age of the learner than any other factor. In this connection, Lenneberg (1967) has argued the most forcefully in favor of a period in the learner's life when language is acquired at an optimum level - 'the critical period hypothesis'.

Lenneberg's arguments stemmed from his observation that children of vastly diverse socio-cultural backgrounds tend to achieve almost the same degree of language competence at corresponding stages of life. This finding, obviously, undermines the importance of environmental factors in favor of the biological factor of age.

As with Penfield & Roberts, Lenneberg relies on the concept of lateralization to substantiate his advocacy of early age exposure to L2. Lenneberg's major contribution to the field, however, may be summed up in his provision of empirical support for the claim that language functions are mostly lateralized in the left hemisphere of the brain.

Lenneberg's (1967) research indicated that injuries to the right hemisphere caused more language problems in children than in adults. More importantly, he found that children were not as prone to language loss after surgeries on their left hemispheres, and were more likely to resume their normal language control after such operations. These findings, obviously, hint at the existence of a direct relationship between age and the delegation of language functions to the left hemisphere, or lateralization.

Later empirical research, however, negated Lenneberg's claim that language learning ability atrophies with the completion of lateralization, which he estimated to occur at or around puberty. For instance, it has been suggested via empirical research that lateralization "appears to be established by age four or earlier" (Krashen, 1973, p. 179).

Be that as it may, it appears that Lenneberg's critical period hypothesis is not too solidly founded on the speculated relationship between language acquisition and brain lateralization. Yet, save this shortcoming in Lenneberg's line of reasoning, his critical age hypothesis may not be substantiated by tying language acquisition to brain lateralization, but instead to some other biological factors which reach their peak around puberty.

2.1.3. Cognitively Based Critical Period
In discarding lateralization as the most prominent factor affecting language acquisition, many theorists have resorted to the principles of cognitive development in their justification of a puberty-bound critical period hypothesis. There seem to be two major pieces of evidence in support of a cognitively-based approach toward the formulation of a critical age hypothesis. First, a multi-purpose use of language seems to emerge simultaneously with many cognitive advances that occur in middle childhood, around the age of seven (Dale, 1972). More characteristically important, however, is the view shared by Krashen (1975), Rosansky (1975), and Felix (1981) that the end of the critical period coincides with Inhelder and Piaget's cognitive stage of formal operations. It is cogently argued that the onset of formal operations, or puberty for that matter, signals a shift in the individual's cognitive development toward the construction of abstract explanatory hypotheses in place of the more concrete ones prior to adolescence. This shift of focus in Krashen's (1975) view may hinder language learning ability since the individual now seeks a conscious understanding of language and in so doing could mentally constrict himself in a labyrinth of rules.

Further support for the delimitation of language acquisition at puberty is provided in Elkind's (1970) assertion that the personality changes typically associated with this stage of life signal the close of the critical period. Moreover, it is maintained by Krashen (1975) that the affective variables which Taylor (1974), along many others have come to link with puberty may, to a large extent, be related to formal operations.

Of still other cognitivistic theories, associative learning may also be discerned as quite accommodating in substantiating the critical period hypothesis and the merits of early age exposure to L2. Proponents of associative learning contend that the individual learns by subconsciously forming an association between the new information and his reservoir of previous knowledge and experiences, and thereby attaches meaning to the new information and makes it more accessible for future use (Stern, 1983).

The process of associative learning is dependent on two major factors. First, for ease of perception, the learner relies on his memory to retrieve the existing information in his cognitive network. Next, for production of the learned material the learner's motor skills should be trusted not to fail him (Steinberg, 1982).

With respect to language perception, Steinberg (1982) expects associative learning to reach its peak around the age of five and to have lost a good deal of its momentum by age ten. This does not imply that learning ceases completely, however, more exposure to language may be necessary to compensate for the loss in the individual's ability to form associations. As for language production, adds Steinberg (1982), quite expectedly, the gradual loss of control over the organs of speech, imposed by age, hampers the older learner's pronunciation. It needs, of course, to be mentioned here that the ability to learn inductively through making associations is one step away from pure rationalism, toward quasi-empiricistic perspectives.

2.1.4. Affective and Social Basis for Age differences
The affective filters related to language learning which come to bear during adolescence were briefly touched upon earlier; however, some theorists have assigned a more pivotal role to them in distinguishing between child and adult language acquisition. In support of the existence of affective and social variables in language acquisition, Smythe et al. (1975) and Taylor (1978) have compared the relative ease with which emigrant children acquire the target language of their newly adopted speech community with the seemingly insurmountable difficulties faced by their parents under the same conditions. It is proposed that emigrant children are more efficient than their parents because they are more adoptive of the new social framework which they encounter. In this connection, Chastain maintains that;

In discussing social and affective factors, however, a distinction needs to be drawn between interpersonal communicative skills and academic, context-reduced, L2 skills. It has been observed that children are better at communicative skills than their adult counterparts for the reasons mentioned above; however, with respect to academic skills, due to their mature cognitive abilities, adults far surpass children (Harley, 1986).

This hypothesis seems compatible with the realities of the L2 teaching environment where teachers have traditionally been content with explanation of abstract grammatical rules. As might be expected, it has been argued by Taylor (1974), that the older learner's advanced cognitive maturity facilitates his conscious understanding of abstract language rules (Harley, 1986). The older learner's cognitive superiority over the younger learner may, according to Ausubel (1964), be traced to his vast repertoire of L1 lexicon on the one hand, and his far greater ability to analytically generalize the grammatical rules of L2 on the other.

Krashen (1977; 1978), however, downplays the role of cognitive maturity by considering it more of a liability than an asset to the older learner's language acquisition. Through his monitor model, he claims that the adult learner's conscious knowledge of the syntactic rules of L2 is detrimental to his production of the forms of language for he constantly monitors his utterances (McLaughlin, 1987). In retrospect, it seems that Krashen's arguments are not at odds with Ausubel's since, as it will be recalled, children and adult learners of L2 may be considered more efficient than one another based on their communicative vis-à-vis academic skills, respectively.

2.1.5. Classroom Interaction and the Nature of the Input
For all intents and purposes, the viewpoints discussed so far may all be stratified along the continuum of planned vs. natural environments. The natural setting affords the learner real life exchanges with native speakers of the target language. As a result, the learner acquires the rules of L2 inductively via his subconscious. Furthermore, in natural settings, the focus is on meaning and exchanges of ideas, or simply stated what language is used for (Chastain, 1988). In planned settings, however, the learner is quite aware of his conscious efforts to learn the target language. Hence, learning takes place deductively, through explication.

In so far as the concept of the 'critical age hypothesis' holds true, it may be argued that the natural setting favors the younger, pre-adolescent learner whose language related faculties are well in tact. However, given the nature of the planned environment, or the classroom, if you will, the older learner, due to his cognitive maturity, not to mention his adjustability to the infra-structure of the classroom, is more successful in learning the rules of L2 than the younger learner.

2.3. Empirical Literature
Lenneberg's theoretical perspectives have served as the backbone for the recent empirical scrutiny afforded to the relationship between the learner's age and his/her language learning ability. Relying on his 'critical period' hypothesis, Lennneberg has strongly favored early age exposure to L2 by asserting that: (a) acquisition in childhood will lead to greater success, and (b) acquisition process will be different between children and older learners (Lenneberg, 1967).

In line with these claims, empirical research has developed in two distinct ways. Firstly, the research guided by Lenneberg's 'Early age exposure=success' assertion has focused on the relationship between the learner's initial age of exposure and his/her overall proficiency in L2. Secondly, his psycholinguistically oriented conclusion has stimulated research on how language is processed at different ages. Hence, the empirical studies reviewed hereunder are categorized into the effects of age on the learner's rate and route of acquisition.

2.3.1. Rate of Acquisition
The parameters traditionally deemed relevant in the rate of L2 acquisition at different ages may be dichotomized into the extent of achievement in: (a) context-reduced or academic skills; and (b) context-embedded or communicative skills. For a more meaningful analysis, the nature of language input is also addressed in terms of planned versus natural exposure.

2.3.1.1. Planned Settings
Ekstrand (1978) in an 18-week program for Swedish school children between the ages of eight and eleven found that older children were better in imitation, listening comprehension, and translation than younger ones. Through the use of the audio-visual method, the role of the teacher was reduced to a minimal level. The results obtained confirmed the existence of a critical period which sets in motion around the beginning of formal operations. This conclusion appears well-founded given the dependence of the audio-visual method on the principles of associative learning. Hence, the older the child, the more likely he is tap into his existing knowledge base to form meaningful associations.

Burstall's (1975) large scale study on 17,000 British school children compared the rate of achievement in learning French of those who began the three-year program at the age of eight with those who started at eleven. Of the skills tested, older learners were found more efficient in listening, reading, and writing, that is, the context-reduced academic achievement measures, whereas the younger learners were more skillful in speaking which may be considered an interpersonal communicative measure.

Olson & Samuel's (1973) study on 100 English L1 participants exposed to phoneme drill treatments of German found that after 10 sessions, adults and junior high students scored higher on pronunciation measures than elementary students. This finding, again, hints at an advantage enjoyed by older learners in context-reduced tasks.

Snow & Hoefunagel-Hohles (1977) research on English speaking participants via audio-tape of five different Dutch words found that in the short term, older learners were more native-like in pronouncing the words; however, the younger learners pronunciation surpassed the adults', in being native-like, as time progressed.

2.3.1.2. Natural Settings
Ekstrand's (1976) study on 2189 English L1 participants, between the ages of 8 and 17, tested on listening comprehension, reading, free writing, pronunciation, and speaking of Swedish L2 found that older learners out-performed the younger ones on these context-reduced measures.

Ervin-Tripp's (1974) study of 31 English speaking participants, aged 4-9, exposed to French for nine months found the older child to be more superior in syntax, morphology, and pronunciation. The tasks were mostly decontextualized and consisted of imitation, translation and acting out of the sentences. It was further inferred that older learners performed more efficiently on the tasks involved due to the 'cognitively demanding' and 'semantically anomalous nature' of the sentences.

In another study involving 1000 immigrant children exposed to English in Canada, Cumming (1981 a) reported that the participants who arrived at older ages performed better on context-reduced skills. However, it was also reported that the time taken to reach a pre-determined level of proficiency was not related to age of arrival.

Fathman's (1975) study on 200 immigrant children exposed to English in the U.S.A. provided further proof for the prediction that performance on context-reduced skills is enhanced with the relative cognitive maturity of older children. More specifically, through the use of such measures as the SLOPE test and picture description he found that older children performed better on morphology and syntax; whereas, the younger participants' performance on pronunciation, solicited by the latter more context-embedded measure, was judged more superior.

2.3.2 Route of Acquisition
In spite of their relative recentness, empirical studies on the effects of age on the route taken in acquisition of language have proved to be phenomenally illuminating. The bulk of these studies have addressed the issue from an L1 perspective, with one more recent exception in L2. However, due to the similarity of the tasks employed in these studies, the conclusions arrived at tend to propagate optimum validity.

Studies on L1 confirm the view that the age factor affects the processing strategy embarked upon in acquisition of language. Fernald & Mazzie's (1991) study compared the strategy sought by the participants in telling a story to toddlers as opposed to adults. It was found that adult participants intuitively resorted to more heavily stressed suprasegmental patterns in relating the story to toddlers than to adults.

The 'Competition Model' of language use has been widely used in studying learners' preferences in sentence interpretation. For instance, MacWhinney & Price (1980), and Bates et al. (1984) discovered that younger children were more likely to follow phonological cues than syntactic ones in their interpretation of sentences. Likewise, Kail's (1989) study on French speaking participants revealed that preference for phonological cues in identifying the agent/subject of the sentences was decisively more rampant in the youngest age category. In a further study addressing the relationship between age and preference for syntactic versus prosodic cues of sentence structure, Read & Schreiber (1982) found that adult university students were less influenced by phonological contours in conflicting sentences than 7-8 year old participants.

Being inspired by studies on L1, Harley (1995) sought to determine whether the same age-related tendencies apply in the processing of L2. Thus, to test the validity of the hypothesis that younger learners in processing L2 follow prosody and older learners pursue syntax, Harley (1995) administered a 30-sentence task to a sample of 56 Cantonese L1 participants and compared the results so obtained with those she collected from a sample of 30 native speakers of English.

The participants were asked to repeat a certain part of each audio-taped sentence that was played for them in sessions lasting about 15 minutes. By repeating the subject noun phrase of the first five sentences as models for the participants to follow, the researcher implicitly directed the participants to repeat this particular grammatical segment of the remaining structures, without mentioning any grammatical terms in her instructions.

The first five model sentences all contained prosodic contours that were in harmony with syntax. The remaining 25 sentences, however, included 10 phonologically ambiguous sentences with conflicting prosodic and syntactic cues. The principle underlying this task was based on the assumption that the participants' preference for prosodic and syntactic cues of the ambiguous sentences is tied to whether they are influenced by the conflicting cues, or if they are able to correctly identify the subject noun phrase of the sentences. Quite naturally, if a participant repeated the prosodically altered segment, in its entirety or parts thereof, it would have to be determined that he was following prosody. By the same token, if a participant could identify the subject noun phrase, without regard to prosodic stimuli, he would have to be following syntax.

Harley (1995) divided her participants into three different age groups, and each group itself was stratified into learners and native speakers. The first age group was composed of 7-to 8-year old participants - 13 learners and 10 native speakers. The second group was made up of 12- to 14- year old participants - 27 learners and 10 native speakers. The last age group of 17- to 23- year old participants included 16 learners and 10 native speakers. All learners were of Chinese descent, with Cantonese L1, enrolled in grades 2, 7-8, and 12 in Canada. None of them had lived in Canada less than a year and some had received formal L2 instruction in their native country.

Harley found out that the learner participants of all three age groups scored surprisingly high scores on following prosody - 5.23, 5.52, and 5.31 out of a maximum of 10 - in their performance on the ambiguous sentences. With respect to syntax, however, their means ranged from 1.92 for age group 1 to 1.00 for age group 2, and 2.69 for age group 3.

The native speaker participants of the first age group scored the highest mean in following prosody, 5.70, with scores of 4.40 and 2.40 registered for age groups 2 and 3, respectively. In following syntax, the older native speakers scored the highest mean, 5.60, followed by 4.30 for the second age group and 1.92 for the first. Based on these results, Harley concluded that her learner participants, in contrast to her native speaker participants, were more influenced by prosody than syntax regardless of their age group. Thus, Harley's (1995) study did not provide enough evidence in support of the hypothesis that younger learners are more attentive to prosody and older learners to syntax.

It needs to be mentioned here that the results obtained by Harley (1995) should not be taken at their face value as sufficient in rejecting the prediction that younger L2 learners differ from the older ones in their language processing. Harley, herself, pointed out that due to the possibility of positive transfer from her learners' highly tonal L1, Cantonese, the validity of her study may have been adversely affected. Furthermore, one may find fault with the non-uniformity of the L2 input that her participants received due to the diversity of the grades from which they were selected. Finally, as a result of a not logically related and sequenced exposure that one receives in a natural environment, the input may not have been optimally comprehensible.

The present study is based on the same hypothesis as Harley's - namely, that younger learners in their interpretation of L2 rely more heavily on prosody, while older learners attend to syntax. However, unlike Harley's sample of L2 learners, the participants of the present study were of a non-tonal L1, Persian. Hence, it is anticipated that the participants of this study would not be as predisposed to phonological cues as Harley's participants were, due to a lack of L1 transfer. Furthermore, the participants of this study were all selected from the same grade and had all received their L2 exposure in a planned environment, thus minimizing the adverse effects of non-uniform and incomprehensible input that Harley's participants allegedly suffered from.

3. Methodology
The method conceived for this study is modeled after Harley's (1995) work. Called the 'Competition Model', it is based on the assumption that the learner's processing preferences come to surface when he/she is presented with conflicting cues at the sentential level. For the sake of optimum comparability with Harley's study, the present study not only replicated the task used by Harley, but also the procedures followed in its administration and analysis of data.

3.1. Participants
The participants consisted of EFL learners of predominantly equi-distant socio-cultural backgrounds enrolled at the Iran Language Institute's level 4, in Shiraz. The deciding factors for choosing level 4 participants, whose proficiency corresponds to ACTFL's low intermediate level, as the target population were that: a) not many young learners were enrolled in higher levels, and b) the topics covered by the sentences of the task are for the most part somewhat child-like and thus not suitable for EFL learners of advanced standing.

At the time of administration of the task, the participants, who were all newcomers to level 4, had been exposed to 220 hours of formal instruction in English in completing the previous levels. Moreover, the participants' grade point average for the previous level fell in the range of 80 to 90. Hence, every effort was made to ensure that the participants were homogenous in terms of their proficiency in L2. Furthermore, it was anticipated that the results obtained from the participants would not be adversely affected by the variable language input due to the uniformity of the syllabi and the teaching method applied to the participants.

The participants consisted of 40 EFL learners of both genders randomly sampled from a target population of approximately 600 level 4 students. The sample was then stratified into the following age groups with 20 participants in each:

a) twelve and thirteen year olds - to comply with Lenneberg's critical age hypothesis which considers this two year span as the cut off period in the individual's language learning ability; and,
b) seventeen and over - to account for the gradual loss of control over the motor skills associated with the adult learner.

3.2. The Task
The participants were asked to listen to a list of 30 audio-taped sentences presented to them individually and one at a time. Following the termination of each sentence the tape was halted so that the participant could repeat the part which he/she identified as the subject noun phrase. The first five sentences contained compatible prosodic and syntactic cues. The tester, himself, demonstrated what was expected from the participant by repeating aloud the subject noun phrases of these five sentences.

The next five sentences also displayed conforming prosodic and syntactic cues to make the participants feel at ease with the task. The remaining 20 sentences, however, contained 10 ambiguous ones with prosodic contours at odds with syntax. To create a sentence with conflicting cues, the prosody for a certain part of a phonologically well-formed sentence was transposed on the corresponding part of another. For example, the prosody for the italicized part of sentence 1 below was used for the underlined part of sentence 2:

Sentence 1: The new teacher's watch has stopped.

Sentence 2: The new teachers watch baseball on television.

Should the participant in repeating the subject noun phrase of sentence 2 not alter the prosody, he/she in all likelihood is more influenced by the prosodic cues presented to him/her. By the same token, if the participant changes prosody to match syntax, he/she may be considered more attentive to the syntactic cues of the sentence.

The conflicting cues of the ambiguous sentences were of three main types. In the first type demonstrated above, the sentence was pronounced in such a way that the stress fell after it normally should (i.e., on watch rather than on teachers). In the second type of ambiguous sentences, the stress fell before its normal position as sentence 4 below demonstrates:

Sentence 3: Our dogs bark at neighborhood cats.
Sentence 4: Our dog's bark sometimes frightens people.

After switching over the italicized segment of sentence 4 with that of sentence 3, in pronouncing sentence 4 the stress is placed prior to the end of the noun phrase.

The third type of ambiguous sentence, although similar to type 2 in the sense that the stress falls before the expected point, is far more challenging since the head noun is not phonologically separated from the rest of the sentence by a pause. The third type of ambiguous sentence is illustrated below:

Sentence 5: When they are wild, animals are brave.
Sentence 6: Trainers of wild animals are brave

A complete list of all 30 sentences with a classification of the more ambiguous ones is provided in the appendices.

3.3. Procedure
The participants were seated individually in front of the tester across a table with two tape recorder/players - one to play the pre-recorded sentences and the other to record the participants' responses. Then, the following instructions were provided to the participants:

4. Results and Discussion
In keeping with Harley's (1995) study, the results obtained were categorized into:

i. performance on unambiguous sentences; and
ii. performance on ambiguous sentences.

4.1. Performance on Unambiguous Sentences
Of the fourteen opportunities they were given to register their preferences, the older learners were successful 11.45 times while the younger learners mean score stood at 10.35. As will be recalled, the unambiguous sentences contained non-conflicting prosodic and syntactic cues. Hence, the participants' scores on items of this type is a measure of their success in correctly identifying the subject noun phrase, or being attentive to syntax.

In spite of the seemingly close performances of the two age groups on the unambiguous sentences, it was found through a t-test that the older group of participants scored a significantly higher mean than the younger one - (see table 1). The said results serve to prove that even in the absence of prosodic stimuli, given the non-conflicting nature of the cues, the younger participants were less apt to follow a syntactically motivated strategy in their interpretation of the unambiguous sentences than the older ones. Hence, it seems that even at this rudimentary stage of the study sufficient evidence is circumstantiated for the hypothesis that younger learners pursue a different route in processing language from older learners.

Table 1 Comparison of Participants performance on unambiguous I Items (Max.=14)

4.2. Performance on Ambiguous Sentences
In comparing the participants' performance on the ambiguous sentences, two platforms offered relevant grounds for analysis. First, a within group comparison compared participants of the same age group on their performance on the ambiguous items by contrasting their preference for prosodic and syntactic cues. Through matched t-tests, it was to be determined whether there existed any meaningful differences between the participants' preference for syntax and prosody within each of the two age groups. Quite clearly, if the hypothesis, that younger learners are more attentive to prosody and older ones to syntax, was to be upheld; it needed to be established that that participants of the younger group would score higher means on following syntax than prosody, while participants of the older group would score higher means on syntax than prosody.

Secondly, a between group analysis of the performance of the participants compared one age group's preferences in interpreting the ambiguous items with those of the other's. Here, a higher score on prosody by the younger group, as opposed to a higher score on syntax by the older group would have provided enough evidence in support of the hypothesis.

4.2.1. Within Group Comparisons
A matched t-test comparison of the participants' mean scores on the strategy sought in interpreting the ambiguous items revealed that the younger group's preference for prosodic cues was significantly higher than their preference for syntactic cues. The older group's preference for syntactic cues over prosodic ones, however, was found to be insignificant (see Table 2).

Table2. Within Group Comparison of Participants Performance on Ambiguous Items (Max.=10)

Key:
Ms= Mean Instances of following syntax
Mp= Mean Instances of following prosody
Significance Level of Ms/Mp Differences(matched t-tests) *p<.01
n.s= Not Significant

With respect to the highly significant difference in the younger age group's preference for prosodic cues over the syntactic ones, it may be asserted that the findings partially verify the hypothesis in the sense that younger learners rely more heavily on prosody than syntax in processing language. However, the older group's non-significant difference does not, without a between group analysis, provide sufficient grounds for the rejection of the hypothesis in absolute terms.

4.2.2. Between Group Comparisons
In order to facilitate this analysis, the data collected was classified into the following response categories, as recommended by Harley (1995):

i. those instances where the respondent followed the sentence syntax in identifying the segment of the sentence to be repeated;
ii. those instances where the participants attended to prosody;
iii. those instances where the syntax of an ambiguous sentence was altered by the respondent to match prosody;
iv. an aggregate of responses to categories ii and iii; and,
v. other types of responses where the respondent either repeated the whole sentence, said one word or nothing at all.

The comparison of responses to the first two categories, via analyses of variance (see Table 3), quite clearly, provided for a most solid foundation for the hypothesis that at the sentential level prosodic cues are preferred by younger learners and syntactic cues win the favor of the older ones. Moreover, a comparison of the two age groups' responses to categories iii and iv lends further support to the hypothesis in the sense that the differences between the means, though not significant, are indicative of a higher amount of attention paid to prosody by the younger age group. In regard to the responses tallied under the last category, the older group's lower mean score suggests that the older participants were less confused by conflicting cues, and hence, more likely to distinguish between syntactic and prosodic stimuli.

Table 3 Between Group Comparison of Participants Performance on Ambiguous Items (Max.=10)

5. Conclusion and Educational Implications
In retrospect, it seems that the overall direction of the findings provides sufficient evidence for the hypothesis that the age factor does play an instrumental role in the route a learner takes in processing L2. As with native speakers of previous studies and unlike Harley's (1995) Cantonese L1 participants, the participants of the present study confirmed the existence of age-related variations in processing L2 for Persian L1 learners of EFL.

In the light of the diametrically opposed findings of the present study and that of Harley's, in spite of the implementation of the same task and procedures, the following explanations may be offered. First, with respect to tonality versus non-tonality of the L1s of Harley's and this study's participants; it may be asserted that her participants were by way of background more biased toward phonology than syntax. In fact, Cantonese is considered one of the world's premier tonal languages; and this may in itself account for the syntactic approaches adopted by Harley's participants across all age groups. In contrast, Persian is not a tonal language, and as a result of a minimal likelihood of positive transfer, the performance of the present study's participants is more reflective of the effect of the age factor than the influence of L1.

Secondly, the participants of the present study were all exposed to L2 in a planned setting. Thus, through such measures as needs analysis, placement tests, and systematic syllabus design, a balance was sought between the learners' existing L2 knowledge and the new material they were to learn. Consequently, the learners were at no time faced with drastically unfamiliar L2 terrain, and this in turn may have resulted in optimum input comprehensibility and the resulting surge needed in the older learners competence to overcome the processing of ambiguous sentences through syntactically-motivated strategies. However, due to the haphazard nature of L2 input in natural surroundings, there is no guarantee that Harley's participants enjoyed the luxury of comprehensible input in their L2 exposure.

Turning now to the pedagogical overtones of the findings, a major effort needs to be undertaken in redesigning those syllabi that do not take age-related needs of L2 learners into account. Clearly, for younger learners, the designers need to place more emphasis on grammar. This, however, would be best accomplished by taking advantage of the younger learners' more acute sound discrimination capacity. For instance, new grammatical points could be introduced and worked on through grammar centered listening comprehension passages and dialogs.

With older learners, owing to their relative weakness in sound perception, the syllabi need to be more focused on phonology. Here, it seems that for maximum efficiency the designers should approach the task by resorting to the older learners' more developed sense of analogy and grammatical competence. It might, for example, be quite a rewarding experience, contrary to general belief amongst practitioners, to lay down the rules governing the phonological component of L2 in clear terms.

Furthermore, in considering the current state of EFL teaching in developing countries, one must not fail to appreciate the importance of homogeneous classes in terms of learners' age. More often than not, because of financial and other extrinsic considerations, learners of various ages are enrolled in the same class with little, if any, attention paid to their needs or varied learning strategies. Learner disorientation resulting from teachers' shifts in focus to accommodate the varied needs of the students is one common outcome of such heterogeneous classes. Teachers, too, often find themselves in an awkward situation in the sense that their attempts to make a particular point understood to a certain age group inadvertently results in another group's alienation.

APPENDIX A

THE COMPETING CUES TASK

List of Sentences:

1. My older sister never cleans her room.
2. The kids in the hall are very noisy.
3. The corner store sells a lot of candy.
4. A cold drink of lemonade tastes good in the summer.
5. The Grade Three class is studying dinosaurs.
6. Most of my friends like school.
7. My new bicycle is broken.
8. Sandy and her friend went to the movies on Saturday.
9. The rabbit in the classroom loves carrots.
10. Too much candy will make you sick.
11. The big black cat had four kittens.
12. The new teachers watch baseball on TV. (Type 1)
13. The fastest runner wins the race.
14. Trainers of wild animals are brave. (Type 3)
15. Your sisters answer questions quickly. (Type 1)
16. My friend Jason came to play yesterday.
17. A large piece of chocolate cake will fill you up.
18. The young student guards don't get tired easily. (Type 2)
19. The girl's lunch was terrible.
20. The leaves on the maple tree are turning red.
21. Only some winter birds fly south. (Type 3)
22. Our dog's bark sometimes frightens people. (Type 2)
23. In winter, windy days are cold.
24. My friends play the piano at school. (Type 1)
25. Our neighbor's fish lives in a tank. (Type 2)
26. His new puppy has run away.
27. Almost all young children like to eat cake. (Type 3)
28. This old house needs to be painted.
29. All my friend's work was lost on the bus. (type 2)
30. Three little kittens lost their mittens.

APPENDIX B
CONSTRUCTION OF AMBIGUOUS SENTENCES

In each pair of the sentences below, the phonology for the italicized part of the first replaced that of the second one:

Type 1 The new teachers watch has stopped.
The new teachers watch baseball on TV.

My friend's play made every one laugh.
My friends play the piano at school.

Your sister's answer surprised everyone.
Your sisters answer questions quickly.

Type 2 Our neighbors fish in the lake on weekends.
Our neighbor's fish lives in a tank.

Our dogs bark at neighborhood cats.
Our dog's bark sometimes frightens people.

All my friends work at McDonald's.
All my friend's work was lost on the bus.

Type 3 When they are wild, animals are brave.
Trainers of wild animals are brave.

At the beginning of winter, birds fly south.
Only some winter birds fly south.

When they are young, children like to eat cake.
Almost all young children like to eat cake.

References
Asher, J & Garcia, G. (1969). The optimal age to learn a foreign language. Modern Language Journal, 311, 334-341.

Ausubel, D.A. (1964). Adults vs. children in second language learning: Psychological considerations. Modern Language Journal, 48, 420-424.

Bates, E., MacWhinney, B., Caselli, C., Devescovi, A., Ntale, F., & Veza, V. (1984). A crosslinguistic study of the development of sentence interpretation strategies. Child Development, 55, 344-354.

Burstall, C. (1975). Factors affecting foreign language learning: A consideration of some related research findings. Language Teaching and Linguistics Abstracts, 8, 5-125.

Chastain, K. (1988). Developing Second Language Skills, Theory and Practice. New York: Harcourt Brace Jovanovich, Publishers.

Cummins, J. (1981a). Age on arrival and immigrant second language learning in Canada: A reassessment. Applied Linguistics, 2, 132-149.

Dulay, H.C., Burt, M.K., & Krashen, S. (1982). Language Two. New York: Oxford University Press.

Dale, P.S. (1972). Language Development. Hindsdale, Illinois: The Dryden Press.

Ekstrand, L. (1976). Age and length of residence as variables related to the adjustment of migrant children, with special reference to second language learning. In G. Nickel (Ed.). Proceedings of the Fourth International Congress of Applied Linguistics, Stuttgart, Hochschulverlag, 3, 179-197.

Ekstrand, L. (1978). English with a book revisited: The effect of age on second language acquisition in formal setting, Didakometry, 60.

Elkind, D. & Lavell, J.H. (1970). Studies in Cognitive Development. Essays in Honor of J. Piaget. New York: Oxford Press.

Ervin-Tripp, S. (1974). Is second language learning like the first? TESOL Quarterly, 8, 111-127.

Fathman, A. (1975). The relationship between age and second language production ability. Language Learning, 25, 245-266.

Felix, S. (1981). On the (in)applicability of the Piagetian thought to language learning. Studies in Second Language Acquisition, 3, 201-220.

Fernald, A. & Mazzie, C. (1991). Prosody and focus in speech to infants and adults. Developmental Psychology, 27, 209-221.

Harley, B. (1995). Second language processing at different ages. Language Learning, 45(1), 43-71.

Kail, M. (1989). Cue validity, cue cost, and processing types in French sentence comprehension. In B. MacWhinney & E. Bates (Eds.), The crosslinguistic study of language processing (pp. 77-117). New York: Cambridge University Press.

Krashen, S. (1973a). Lateralization, language learning, and critical period. Langauge Learning, 23, 63-74.

Krashen,S. (1975). The critical period for language acquisition and its possible bases. In D. Aaronson and R. Reiber (Eds), Developmental psycholinguistics and communicative disorders (pp. 211-224). New York: Newy York Academy of Sciences.

Krashen, S. (1978b). The monitor model for adult second language performance. In M. Burt, H. Dulay, and M. Finocchiaro (Eds.), Viewpoints on English as a Second Language (pp. 152-161). New York: Regents.

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

Lenneberg, E. H. (1967). Biological foundations of language. New York: Wiley.

MacWhinney, B., & Price, D. (1980). The development of the comprehension of topic-comment marking. In D. Ingram, C.C.
Peng, & P. Dale ( Eds.), Proceedings of the First International Congress for the Study of Child Language. Lanham: University Press of America.

Olson, L., & Samuels, S. J. (1983). The relationship between age and accuracy of foreign language pronunciation. Journal of Educational Research, 5, 261-285.

Rosansky, E. (1975). Methods and morphemes in second language acquisition. Language Learning, 26, 409-425.

Seliger, H. ,Krashen, S., & Ladefoged, P (1975). Maturational constraints in the acquisition of a native-like accent in second language learning. Language Sciences, 36, 20-22.

Smythe, P.C., Stennett, R. G., and Gardner, R.C. (1975). The best age for foreign language training: Issues, options and facts. Canadian Modern Language Review, 32, 10-23.

Snow, C., & Hoefnagel-Höhle. (1977). Age differences in the pronunciation of foreign sounds. Language and Speech, 20, 357-365.

Taylor, D. M., & Jaggi, V. (1974). Ethnocentrism and causal attribution within a south Asian context. Journal of Social Psychology, 5, 162-174.