Spanish Speakers Semi- and Residually Native: After the Placement Test is Over

University of Texas, El Paso

The sociolinguistic realities of a varyingly bilingual international metropolitan area such as El Paso, Texas/Ciudad Juárez, Chihuahua are complex¹²⁸, and this complexity is inevitably reflected in the linguistic profiles of students enrolling in Spanish courses at a large (ca. 15,000), essentially open-admission and overwhelmingly commuter-student institution such as the University of Texas at El Paso (UTEP), whose enrollment patterns (ca. 55% Mexican-American or Mexican, 4% Black, about 37% U. S.-educated «other», and the rest variously non-hispanophone foreign) approach north-of-the-border population percentages (about two-thirds of El Paso's residents are Hispanic), and which actively recruits in Mexico.

Since 1972, the UTEP Department of Languages and Linguistics' four lower-division (Frosh and Sophomore) Spanish courses have been divided into two tracks, one for «Native Speakers» and the other for «Non-Native Speakers». I encircle these designators in quotes to reflect the area's sociolinguistic realities, which can only be described in terms of a continuum ranging from total monolingualism in English to total monolingualism in Spanish and all points in between, though as is true of continua, clusters can be found, the most noteworthy of which involve Mexican-Americans, typically prompted by circumstances to be bilingual in English and Spanish to a greater or lesser extent. In a best of all possible worlds, lower-division Spanish at UTEP would offer as many tracks as there were continuum clusters or archetypes of students¹²⁹. In the real world of scheduling problems, minimal textbook variety, ignorance or even hostility among staff and extra-departmental faculty, and widespread student antipathy or indifference toward a language requirement, even a two track system is a constant challenge to operate as conceived. While Klee and Rogers (1989) substantiate by surveying a wide variety of undergraduate programs across the nation what the profession has long intuited to be true, namely, that the placement problem «mentioned most frequently by the respondents was that of false beginners», at UTEP the central concern is the native or near-native speaker who seeks to enroll in Spanish One for Non-Natives (4101). A month after I assumed direction of Spanish placement in 1978, I estimated that about 40% of the students in my own section of 4101 were native or quasi native hispanophones; requests that other instructors of 4101 come up with their own estimates produced identical results. It goes without saying that such mixed classes characteristically produce two sets of attitudes: cynicism among natives or near-natives, who correctly assume they need never crack a book to get at least a B (and who thus learn nothing from the course), and high anxiety among the non-native or largely so who, in addition to experiencing all the fears of foreign language so carefully catalogued by Horwitz 1989 and elsewhere, must also frequently suffer the barely-concealed mirth of hispanophone peers.

Since 1979, our department has been perfecting a Spanish placement system which as of this writing (July 1989) represents, we feel, the best we can devise. It works like this: no student is to be allowed to register for the first time for a college-level Spanish course without taking the placement test (a 100-item multiple-choice machine-gradable local instrument) or signing what we call a waiver, i. e., one of two sternly-worded documents which allow students to affirm that they do not speak Spanish at home or in the neighborhood and did not recently study it for two consecutive years or more in high school¹³⁰, or to affirm that though they speak Spanish at home/in neighborhood they have not acquired literacy skills in the language. Without exception, any student not qualified to sign a waiver -and any student requesting a waiver must answer a series of questions about which more later- must take the Spanish placement test. In

––––––––   817   ––––––––

any given year, about 1,000 students take the test, ca. 550 for the fall semester, 350 for the spring semester, and 100 for the summer term. An equally large number sign waivers in identical proportions but disproportionately according to intended course; thus from November 14, 1988 through January 23, 1989, i. e., before, during and after spring registration, 354 persons signed waivers of which 282 were for Spanish 4101 (non-natives) and 72 for Spanish 4103 (native speakers).

Students taking the Spanish placement exam -and 323 did so before/during/after Spring 1989 registration- are not allowed to leave the exam room or the immediate registration area until they have received their score sheet. We insist that students receive their test results upon completion because in the past at least a quarter of all examinants would not return on a subsequent day to learn their scores and would then claim at registration that because they «couldn't get» the results («The office was closed when I got there», «The secretary didn't have the computer printout», «I had to work late that day and couldn't phone», «Nobody told me I had to come back»; and a dozen other equally inventive tales), they were entitled to take Spanish 4101 after all.

Students approach the exam administrator's desk individually and must sit alongside the desk while the administrator hands over the score sheet (a copy of which the department permanently archives) and explains its contents. One disadvantage of a two-track system is that many students do not readily grasp the relationship between course numbers and tracks and levels, i. e., that Spanish One is 4101 if for non-natives but 4103 if for natives, that Spanish Two is 4102 if for non natives and 4104 if for natives, etc., and this must often be explained with care. Students are also told how many credit hours they will receive upon successful completion of the course they place into. (For one year in the early 1980s we granted automatic course credit for a satisfactory exam score alone. We were so overwhelmed with examinants that we discontinued the policy). In addition they are told to present the score sheet at registration and then again to the departmental secretary once they have passed the course.

The Spanish placement exam has been normed before on several occasions, but until recently no attempt has been made to determine exactly what percentage of examinants constituted the population which I and my colleagues must typically interview before deciding what course to place them in, though I have known for years intuitively that students who score as «borderline» on the ten native speaker identification questions (NS IDs) are the ones for whom the score alone does not suffice. The ten NS IDs all involve colloquial words or expressions that are characteristic of Mexican-origin Spanish and that seldom if ever make their way into Spanish-as-foreign or-second-language textbooks. Repeated ad ministration of these items -always embedded in others so as to avoid undue focus- has repeatedly shown them to be the single most reliable component of our placement test: students that instructors know to be thoroughly anglomonomatriphonic at the start of Spanish 4101 invariably score zero to -rarely- three correct on the ten NS IDs while students we know to be fully and natively fluent always score from seven (usually eight) to ten correct. The problem is the existence of the above-referred-to continuum: not all are truly non-native or fully native.

Of the 323 students taking the Spanish placement test for Spring 1989, the clear majority (242/323 = 74.92%) were tracked by their answers to the NS IDs as native speakers and placed into any one of the four Spanish or Native Speakers courses (I = 4103, II = 4104, III = 3203, N = 3204) or in a small number of instances directly into any third year literature, linguistics or composition course. The following table (Table One) provides raw numbers and percentages and also indicates the number of correct answers given on the ten NS IDs per examinants placing into each course.

As Table One demonstrates, the lower the course level the lower the mean score on the ten NS IDs. Thus the mean NS ID score for students placing into 33xx was 10, for 3204 9.90, for 3203 9.48, for 4104 8.91, and for 4103 7.03. Note that a considerable drop-off occurs between 4104 and 4103. While only three of the 80 examinants who placed into 4104 (and none of the 3203/3204/33xx examinants) scored as borderline speakers (i. e., received NS ID scores lower than seven but higher than three), 13 of the 40 -nearly one third- of those placed into 4103 did so.

Analysis of examinants tracked as non native speakers (Table Two) is particularly illustrative of the difficulties encountered in attempting to adequately place certain population clusters on our area's language continuum.

Particular attention is drawn to those students who placed into 3201 (Spanish Three for Non Natives), as 33xx¹³¹ and 3202 produced only one examinant each, and none of the 13 examinants placed into 4101 tracked as borderline native speakers. Of the 20 students the exam placed into 3201, nearly half (9/20) tracked as borderline native speakers with a mean NS ID score of 4.67, and while a smaller percentage (18/48 = 37.50%) of the examinants who placed into 4102 tracked as borderline natives, their mean NS ID score at 4.89 is very close to (though a bit anomalous as slightly higher than) the mean NS ID score of the nine borderline-tracked examinants whose higher overall scores placed them into Spanish Three for Non-Natives (3201). The mean NS ID scores of the eleven examinants tracked as non-natives and placed into 3201 -1.91- and the mean NS ID scores of the 28 tracked as non-natives and placed into the next lowest course 4102 -1.29- are considerably more in keeping with what we would expect in a non-native track and for the course levels indicated, and represent a gradual decline from an already low high. (Note that the mean NS ID scores of all 13 examinants placed into 4101 -0.92- also fits this pattern). It is the nine examinants tracking as borderline and placing into 3201 plus the 18 examinants similarly tracking but placing into 4102 who constitute -together with the 13 examinants who tracked as borderline and were placed into 4103 (see Table One)- an aggregation of students whose ultimate placement assignment must be made with the greatest of care, and only after the examinant has been interviewed as extensively as exam room conditions permit. Further support for the need to question these students carefully so that an adjustment in ultimate course placement can be made if necessary -from the native to the non-native track or vice-versa, or from a higher level to a lower one (though never the reverse)- derives from a consideration of the total scores (= x items correct out of 100 items total) of the six sub-groups (4103 borderline and native, 3201 borderline and non-native, 4102 borderline and non-native). Table Three gives these, along with subsequent comparisons with groups whose potential to be ill-served by the placement test is incontrovertibly low.

As Table Three shows, while the student population placed into Spanish 4103 (native speakers) is essentially homogeneous even though a third of this group tracked as border line (see Table One), the mean total placement exam score -51.22- of those students tracked as borderline and placed into Spanish 3201 (Sp. Three for Non-Natives) is only 4.01 points lower than the 55.23 mean score achieved by Sp. 4103 borderlines. Quite simply, these two groups are too close for comfort. Fortunately just nine examinants tracking as borderlines were placed into Sp. 3201 for Spring 1989. My general policy is to allow such high placement (and its attendant eleven credit hours, eight for the exam and three for the course) only to students who have previously indicated, in response to questions I ask them before I determined their placement and show them their score sheet, that they have recently studied at least two consecutive years of Spanish in high school and that they did well in the course. (More on these and other interviewing tactics below). While the gap between Sp. 3201 borderline and Sp. 3201 non-native placees is not narrow (at 3.49), it is considerably narrower than the one separating Sp. 4102 borderline and Sp. 4102 non native placees -a full 11.19 points. I have undertaken no investigation of how borderline-tracked examinants who placed into Sp. 4102 fare as compared to both non-natively tracked 4102 placees and to the clear majority of 4102 students -roughly 75% of them- who placed or more typically waivered into 4101 and have enrolled in 4102 because they passed 4101, the antecedent course, and, indeed, would be very reluctant to undertake such investigation because of the large number of variables involved. But it is my intuition as coordinator of lower-division Spanish and frequent instructor of 4102 that it, more than any other course, is where we encounter the largest number of students who have fallen through the cracks or for whom the dichotomous two-track «native vs. non-native» system does not and cannot work. I taught a section of 4102 during Spring 1989, and the encapsulated language-background biographies and classroom performance stories of two of my students, Larry and Dan (not their real names), will illustrate our system and how, with students not neatly classified by its placement exam component, it sometimes works and sometimes does not. The following narratives represent only two of many different personal stories.

Larry is the product of an ethnically-mixed marriage, a phenomenon not uncommon in El Paso. His Mexican mother has spoken «some» Spanish to him since infancy, but the predominant home language is English, and the family lives in a lower-middle class neighborhood where Hispanics are about a third of the population. On the placement test, Larry scored as follows: NS ID 5 (i. e., borderline), and 38/100 total. Following about five minutes of questioning I decided that 4102 (Sp. Two for Non-Natives) was the best if not the ideal choice. Except for «a year or so» of required Spanish in grade school, Larry had never studied the language. Larry's initial attitude was one of nonchalance: it was clear that he was not preparing for class and expected to get by -even do well- on the strength of his home-acquired language ability, though two successive low exam grades convinced him otherwise. He ended up with a C in the course. On the monolingual-to-monolingual continuum I would classify Larry as, at best, a residual native speaker of Spanish, with a number of home-acquired oral-aural skills to be sure, but definitely closer to English monolingualism than to its Spanish counter part. Spanish 4102 was appropriate for Larry in that it taught him that college Spanish is serious business, that residual acquaintance would not suffice to get him through, and that the teacher and the textbook could indeed contribute to expanding his command of the language.

Dan's story is less encouraging, just as his language background configuration is more complex. The product of two Mexican-American parents and a lower-middle-to-working class neighborhood that is about 90% Hispanic, Dan reported that while his parents, aunts and uncles spoke mostly Spanish to each

––––––––   820   ––––––––

other, they mainly used English with him, and he, in turn, spoke only English to peers and siblings, but recalled that he had spoken only Spanish to his grandparents when he was younger. He also reported using «some» Spanish at work and with a few neighbors. One of several thousand students to be processed during the three hectic days of Fall 1988 registration, Dan signed a waiver for 4103 (One for Natives) because he more or less considered himself hispanophone. However, in 4103 he did poorly -his course grade was a D, and he claimed he never understood what the instructor was saying (a judgment his former instructor confirmed to me)- and came to me requesting that he be allowed to switch tracks and take Spanish Two for Non Natives. Following extensive interviewing I agreed, though reluctantly, for I found Dan to be a true borderline: not non-native but not native either. Dan's performance in my section of 4102 confirmed my fears. Though Dan never used his strength in spoken Spanish as grounds from which to ridicule his classmates, it was clear they were aware from the outset that here was someone whose proficiency was not classroom based. Dan always did well in communicative exercises and consistently failed to get the point of structural drills; his capacity for comprehending deductive explanations and understanding metalanguage was practically non-existent; he was only beginning to understand that there was a formal (let alone a functional) difference between subjunctive and indicative when the semester ended. Yet his home-based acquisition sufficed to earn him a B in the course. The implications of Dads story are two-fold: that a certain number of truly borderline students will never be well served by either component of a dichotomous lower-division Spanish program, and that in a course which emphasizes communicative over structural activities, the Dan-like student will succeed in doing fairly well without expending much effort.

This second issue is beyond the scope of the present paper so we return to our primary concern: granted that a certain number of Dans can never be properly placed in a dichotomous program, what can be done nonetheless to assure that we know which type of student runs the greatest risk of being misplaced, and that we know him to interview potentially-misplaced students so they are put into whichever course will prove the least injurious and the most beneficial?

From Tables One, Two and Three and our previous discussions of them, we already know the nature and size of the group and groups at greatest risk: 4103, 4102 and 3201 borderlines. In the spring of 1989 there were 13 of the first, 18 of the second and nine of the third for a total of 40 examinants out of 323 persons taking that semester's test or 12.39% of the total. To that number must be added the students who sign not only waivers for 4103 (which 72 students did for Spring 1989) but for 4101 as well (an impressive 282 for the semester). The reason one must inter view students requesting waivers is obvious: most have given no thought to their back ground in Spanish and are even unaware that our program offers two separate tracks let alone what the difference is between them. (Some students are even unaware they cannot start their work at a level higher than first semester without taking the placement test to prove they are capable of doing so. And some who do know their background in Spanish are only too eager to hide it in hopes they can take a course they correctly assume will exempt them from work). Thus one out of every eight placement exam takers and all waiver signers must be interviewed. In Spring 1989 that added up to 394 individuals equalling 58.20% of all students seeking or contemplating enrollment in lower-division Spanish.

The easiest groups to interview are those students requesting waivers for Spanish One for Non-Natives (4101), or students who do not know which waiver they want or if they want a waiver at all. One asks if the student has taken Spanish before and if so for how long and where and how recently. One also asks whether the student is spoken to and/or speaks Spanish at home. Demonstrably non-native students with fewer than two years of recent high school Spanish are typically allowed to read the 4101 waiver with no further ado and sign it if appropriate. Here the questioning turns delicate, for under no circumstances must one give the impression that one assumes the student is hispanophone because of surname or physical appearance and in any event, surname and appearance are often poor guides to language background. Keep the questions linguistic and clear. Never ask, «Do you speak Spanish at home?», since this «you» is often misinterpreted as a marker of middle voice, i. e., «Is Spanish spoken in

––––––––   821   ––––––––

your home?» Ask, instead, «Do you speak Spanish to your parents?» (subject, verb and prepositional emphasis deliberate). If the answer is yes, then corroborate it by asking to whom else the student habitually speaks Spanish. In typical Likert Scale fashion, someone who speaks Spanish to parents is also likely to do so to parents' peers be they relatives or friends. At that point it can be safely assumed that the student is a native speaker, so the next question is: «Do you read Spanish?» If so, the student is required to take the placement test. Native-speaking students who insist they do not read Spanish are given the 4103 waiver to sign. Similar questions about reading (and writing) Spanish are routinely asked of students who request the 4103 waiver to begin with. I estimate that about half this group ends up taking the placement test. In general it is a good idea to insist that any student with whatever degree of home based Spanish background take the placement test, especially when we learn that of the 354 persons signing waivers for Spring 1989, 308 did so during the three days of on-site registration conducted amid standard bedlam in the halls of UTEP's noisy Special Events Center, where lines are long, students confused, and faculty irritated. Under such circumstances, a few hurried questions often fail to achieve their goal.

During the third week of classes in Spring 1989 I requested that the instructors of all three sections of Spanish One for Natives (4103) and all nine sections of Spanish One for Non-Natives (4101) administer a three-page questionnaire to all students attending that day. Information requested included whether students signed waivers or took the placement test to gain entrance to the course, their general schooling background (i. e., any bilingual education? any schooling in Mexico?), their present-day and early-childhood language usage patterns, and their previous study of Spanish at school. I found that students in Spanish 4101 overwhelmingly be longed there. At most ten percent of them appeared to be semi- or residually native speakers of Spanish (and this despite the fact that 22.22% of them had Spanish surnames), and I recommended in my report to colleagues that given our program's dichotomous nature it is inevitable and even necessary that some semi- and most residual natives end up in the non-native track, as every semester I am visited by from ten to fifteen students enrolled in Spanish One for Natives (4103) who insist they cannot understand their instructor's Spanish and who through further testing reveal their spoken skills to be markedly weaker than the typical north-of-the-border native speaker with whom they must compete in class. Students in Spanish 4103 presented a more complex picture, one probably deriving in no small part from the fact that 75% of them signed waivers to enter the course. I concluded that about 12% of that semester's 4103 enrollees might have been able to place into 4104 (Two for Natives) had they taken the test, inasmuch as 12.19% indicated they spoke «only Spanish to one or both» parents and that an equal percentage indicated they had studied Spanish in high school for two years (during which they had surely picked up literacy skills). At the other end of the spectrum, perhaps eight to ten percent of the semester's 4103 enrollees may be at best semi- or even residual natives and might be better served by 4101 (or the catch-all 4102); thus seven students (8.54%) indicated they spoke no Spanish to either of their parents and six said they spoke no Spanish to older relatives.

Many of the questions typically asked of students requesting waivers are also asked of that 12% of placement test takers who track as borderlines. But particular attention is paid to the examinants previous and current language-use patterns. In addition to the afore mentioned questions about languages spoken to and received from parents, information is sought about language usage at work, with peers, with siblings, with neighbors, and with relatives both close and distant, including number of contact hours with same during a typical week. Finally, and especially if the usage picture is mixed or cloudy, it is often helpful just to ask: «Do you consider yourself a speaker of Spanish?» This question is disarming, in particular as the capstone to a long series of admittedly personal questions which leave some respondents slightly unnerved.

It is also crucial to pay close attention to educational background specifics. In cities such as El Paso where there is no cross-district busing, to know what high school your examinants attended is to know their neighborhood -and it is extremely important that placement test administrators be knowledgeable about prevailing language-use patterns

––––––––   822   ––––––––

throughout the region their university serves. Answers must also be sought to questions about the amount of contact the borderline examinant has had with a monolingual hispanophone country (invariably Mexico in UTEP's case): How frequently do you cross over to Cd. Juárez and why? How frequently do you visit friends in Juárez or elsewhere in Mexico? How frequently do they visit you? Have you traveled into the interior of Mexico? Have you resided there extensively?

When responses to all these questions still produce a mixed or typically neither/nor picture, and if there is time, I invite the examinant to read out loud a passage in Spanish. Results from this -which would take another paper to discuss- are quite revealing, considerably more so than free conversation in Spanish, which offers too many opportunities for the respondent to engage in linguistic avoidance behavior (which in itself is indicative but quite time-consuming). If even a reading passage fails to clarify the examinant's status as native or non-native, it is then and only then that I ask: «Which course would you prefer: X or Y?» I estimate that I am reduced to this admission of diagnostic inadequacy in about ten percent of all extensive interviews.

In sum, for one out of eight persons taking the placement exam and for an estimated 25% of all students requesting waivers, interviews ranging from the middling to the maximal are an absolute necessity for all the reasons I cite. In the land of the continuum -and where bilingualism abounds, continua prevail- the one-eyed (or one-languaged) is if not the exception then hardly the norm, and provision must be made for the linguistically conjunctivitic.

Horwitz, Elaine K, «Facing the Blackboard: Student Perceptions of Language Learning and the Language Classroom». ADFL Bulletin 20:3 (April 1989): 61-64.

Klee, Carol A. and Elizabeth S. Rogers. «Status of Articulation: Placement, Advanced Placement Credit, and Course Options», Hispania 72 (1989): 264-74.

Ornstein, Jacob, Guadalupe Valdés-Fallis and Betty Lou Dubois. «Bilingual Child-Language Acquisition Along the United States-Mexico Border: The El Paso-Ciudad Juárez-Las Cruces Triangle». Word 27: 1-3 (April-August-December 1971-1976) (Special Issue on «Child Language/1975»): 386-4114.

Teschner, Richard V. «Historical-Psychological Investigations as Complements to Sociolinguistic Studies in Relational Bilingualism: Two Mexican-American Cases». The Bilingual Review/ La Revista Bilingüe 8 (1981): 42-55.

_____. «Second-Language Acquisition and Foreign Language Teaching: Spanish Language Programs at a University on the U. S.-Mexican Border». In Bilingualism and Language Contact: Spanish, English, and Native American Languages, ed. by Florence Barkin et al., New York, NY: Teachers College Press, 1982. 228-40.

_____. «Spanish Placement for Native Speakers, Nonnative Speakers, and Others». ADFL Bulletin 14:3 (March 1983): 37-42.

Valdés, Guadalupe. «Bilingualism in a Mexican Border City: A Research Agenda». In Bilingualism and Language Contact: Spanish, English, and Native American Languages, ed. by Florence Barkin et al., New York, NY: Teachers College Press, 1982. 3-17.

_____. «Teaching Spanish to Hispanic Bilinguals: A Look at Oral Proficiency Testing and the Proficiency Movement» Hispania 72 (1989): 392-401.

Hispania [Publicaciones periódicas]. Volume 73, Number 3, September 1990