The Acquisition of the Spanish Prepositions Por and Para in a Classroom Setting
Author(s): Derrin Pinto and Scott Rex
Reviewed work(s):
Source: Hispania, Vol. 89, No. 3 (Sep., 2006), pp. 611-622
Published by: American Association of Teachers of Spanish and Portuguese
Stable URL: http://www.jstor.org/stable/20063366 .
Accessed: 23/01/2012 13:08
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
American Association of Teachers of Spanish and Portuguese is collaborating with JSTOR to digitize, preserve
and extend access to Hispania.
http://www.jstor.org
Applied Linguistics
Prepared by Joseph Collentine
The Acquisition
of the Spanish Prepositions/wr
in a Classroom
Setting
and/jara
Derrin Pinto
of Saint Thomas
University
Scott Rex
Southern Oregon
University
the prepositions
por and para using a corpus that includes
study, we analyze
learners of L2 Spanish. The main objective
of this analysis
is to investigate
the ac
for learners who have only studied in a classroom
curacy
setting and who have not studied
or lived abroad in the target culture. In light of the findings, we then consider current strategies for teaching por
common
and para
and note ways
in which
the most
found in textbooks,
the tendency
explanations
especially
to present
them in opposition,
coincide with
the difficulties
learners have or with
the pre
may not necessarily
In this cross-sectional
Abstract:
four
of undergraduate
of these prepositions
levels
positions'
Key Words:
1.0
functions
in discourse.
classroom
Introduction
learning,
L2
Spanish,
prepositions,
por
and para,
second
language
acquisition
and Goals
prepositions por and para (P&P) in Spanish represent an established grammar point
that appears in virtually all second- and foreign-language
(L2) texts from beginning com
texts to advanced grammar and composition books. These two prepositions,
municative
which inmany cases correspond to "for" inEnglish, are traditionally presented in opposition and
not in conjunction with other prepositions. The repeated attention dedicated to P&P is a conse
The
quence of the difficulty in acquiring their various functions, and such extensive consideration in
the classroom would lead us to expect some measurable gains in accuracy. However, according
to our research, learners show modest progress in the acquisition of P&P over a four-year
sequence of courses.
studies on P&P acqui
Surprisingly, there exist no published longitudinal or cross-sectional
sition in a classroom setting. The two available studies to date on por and para (Guntermann
1992; Lafford and Ryan 1995) investigate the acquisition of these prepositions using a total of
eighteen informants who have spent time studying or living abroad in the target culture. Gunter
university
mann's
data reveal the expected progression from extremely inaccurate use of por and para at the
novice level (0% with only one attempt) to a high accuracy rate of 94% at the superior level. Her
data also reveal a slight decrease in accuracy (from 72% to 70%) between the Advanced and
Advanced Plus level. Although not statistically significant inGuntermann's
study, this decrease
in accuracy at the intermediate level appears to be a common phenomenon
in the acquisition of
P&P.
Lafford and Ryan's data also reveal a "U-shaped" acquisition curve (see Kellerman
1983 for
a discussion of this pattern), inwhich the decrease in accuracy at the intermediate mid-level
is
particularly pronounced wither,
dropping from 67% at the novice mid-level to only 41% at the
In addition, Lafford and Ryan propose a tentative order of acquisition of
intermediate mid-level.
P&P and examine the semantic functions of the prepositions in the interlanguage of the students.
Their data reveal that the most common nonnative-like uses are P&P as substitutions for other
prepositions or conjunctions and P&P in contexts inwhich nothing is required in Spanish.
"The Acquisition
of
the Spanish
Prepositions
por
and Scott Rex
Pinto, Derrin
in a Classroom
and para
Setting"
89.3 (2006):
611-622
Hispania
H?spanla 89 September 2006
612
Padilla-Falto (1996) in an unpublished dissertation investigates learners in the classroom en
vironment; however, all the subjects are at the advanced level. These learners were divided into
three groups: one group received grammar-based instruction of P&P in a teacher-centered class
room
a second
environment;
group
went
through
a
meaning-based,
learner-centered
treatment;
the third was a control group that received no instruction on the prepositions. The findings
showed that both experimental groups performed better than the control group, but the effect
was not long-term. Echoing the results of Guntermann
(1992) and Lafford and Ryan (1995),
Padilla-Falto also found that the majority of errors were not due to confusion between por and
an important issue towhich we
In another dissertation, Yee-Wun
two different treatments of P&P. The
would find in a normal, second-year
para,
return below.
Lam (2003) exposed two groups of second-year learners to
treatments were limited to a 45-minute lesson, much like one
language class. Each consisted of a grammar explanation in
will
of the meanings of P&P, examples of usage, and practice with both comprehension and
production. The treatments differed only in the manner in which the grammar was presented.
While one group received an explanation of the interrelated meanings of P&P, the other group
was presented only with a list of the various uses of each preposition. Post tests revealed no
statistically significant differences between the two groups. In section 4.0, we will discuss the
English
of P&P in such a short period of time and
inefficiency of presenting the multiple meanings
to
the
other
ways
topic.
approach
suggest
The two abroad studies indicate that L2 learners, when immersed in the target culture, show
a fairly predictable pattern of acquisition of por andpara when correlated with oral proficiency.
research shows immediate gains for advanced students who have been exposed
Padilla-Falto's
to some form of P&P treatment. Missing are data collected from lower and intermediate learners to
levels. Padilla-Falto mentions that "future research
investigate acquisition across developmental
should look at the frequency of errors at different levels of proficiency" (197). Our study sets out
to determine how the typical four-year classroom experience, in the absence of a study-abroad
stay, impacts the acquisition of por andpara. We address the following research questions:
1. Do L2 Spanish
tions
2.
3.
4.
over
learners show an increase in accuracy with
a four-year
university
these problematic
preposi
experience?
What are the most common accurate and inaccurate uses of P&P at each level?
courses exhibit the same type of gains that Guntermann
Do students in upper-division
Lafford
and
and
Ryan (1995) found with their advanced learners who had spent
(1992)
time abroad?
How well do current L2 textbooks address the difficulties that learners face in acquiring
the multiple uses of P&P?
In addition to addressing these questions, we also discuss the implications that our findings
have for teaching and consider the most problematic uses of P&P that impede progress in the
acquisition of these prepositions throughout traditional university coursework in Spanish. Ul
timately, we question the effectiveness of the customary teaching of P&P in L2 classes and make
this approach to bring teaching of P&P in line with our current
for modifying
recommendations
understanding
of the general acquisition
pattern followed
by most
students.
2.0 Methodology
2.1 Participants
learners of Spanish at the University of California at Davis
A total of 80 English-speaking
our
A
in
background questionnaire was administered to facilitate the
study.
(UCD) participated
elimination of all native and heritage speakers and all students who had spent more than two
weeks in a Spanish-speaking
country. Four groups of 20 students were tested at four stages of
Acquisition
of Por
and Para
613
study, roughly representing the four years of undergraduate study in Spanish atUCD. The series
of Spanish classes atUCD, which adheres to a ten-week quarter system, is as follows: Spanish 1,
After Spanish 100, which is an introduction-to-literature
course, students
2,3,21,22,23,24,100.
take various upper-division
literature, culture, and linguistics courses. The four stages chosen
for this study are Spanish 3,22,24,
and an advanced group comprised of students from 100 and
courses. Each of the first three levels represents the endpoint of a se
various upper-division
quence using a particular textbook, while the fourth level represents a random sample of
students, primarily
course
division
Spanish majors
and minors,
who
have
successfully
advanced
to upper
work.
The strict organization of the language program atUCD provides an exceptional opportunity
to study the impact of classroom instruction on the acquisition of P&P. All incoming teaching
assistants take a methods course and are trained in communicative
language teaching. Each of
the three lower-division
sequences is supervised by a faculty coordinator who regularly ob
serves the instructors, and a standard syllabus and testing materials are used for all courses. The
following is a brief explanation of themethodology
the data were collected:
1.
2.
1 (Spanish 3): The first-year sequence adheres to a highly communicative
ap
proach. The textbookDos mundos (Terrell, Andrade, Egasse andMu?oz 1998) was used
for the three first-year courses. Dos mundos is based on the Natural Approach and
claims that communication
is the core of their program. Each chapter is loosely centered
on a topic (classes, family, personal information, plans and preferences, activities and
Level
points follow a fairly typical order. In this book, P&P appear
chapter covered in this course. Since the students completed
in the semester, they had already completed this chapter.
second-level sequence is a continuation of the first-year pro
gram. While still emphasizing oral proficiency, this level includes an increased focus on
the development of reading and writing skills. This intermediate level used A lcorriente
(Blake, Ramos and Marks 1998) when the data were collected. Many basic grammar
points and verb tenses are covered (or reviewed), often inmore detail than in an intro
places, etc.), and grammar
in Chapter 10, the second
the test for this study late
Level 2 (Spanish 22): The
ductory
3.
4.
and texts thatwere utilized for each level when
text,
and
some
cover
more
advanced
structures
(relative
clauses,
sequence
of
tenses, etc.). The presentation of P&P in this text occurs in Chapter 10, which is the
second chapter covered in the course. Thus, students from this level had also been
exposed to the text presentation prior to participating in this study.
Level 3 (Spanish 24): The third level is designed to prepare students for upper-division
coursework. The focus is on composition, primarily in an academic register. Palabra
abierta (Colombi, Pellettieri and Rodriguez 2001 )was the text used for this level. Gram
matical structures are reviewed in a framework of functional grammar and in the context
of formal academic writing. Short readings are also included. This text does not contain
any explanation of P&P, although they do appear in the some of the expressions.
Level 4 (Spanish 100 and upper-division
classes): Spanish 100 is an introduction-to
literature course, which serves as the gateway for all students with a major or minor
before they go on to take upper-division
courses in literature, culture, and linguistics.
For this study we also collected baseline data from 44 native-speakers
using the same instrument (discussed below).
(NS) from Spain and
Mexico
2.2
Instrument
The instrument used, referred to as a Discourse Completion Test (DCT), was originally de
signed for eliciting a range of speech acts as part of a study in interlanguage pragmatics and was
not originally created to target specific grammatical structures or lexical items. The DCT is a
614
89 September 2006
Hispania
written questionnaire that includes brief descriptions of situations followed by blank spaces to
be filled inwith the subject's own version of an appropriate response. The objective behind as
sembling this 20-item DCT was to choose a representative sample of speech acts that covered an
extensive range of language functions and difficulty. The speech acts included are the following:
requests, expressions of gratitude, complaints, compliments, compliment responses, apologies,
refusals, invitations, accepting invitations, greetings, and farewells. The DCT was administered
over the Internet in a supervised setting during the last two weeks of the ten-week quarter.
The following is a sample scenario from the DCT that elicited expressions of gratitude: "Des
pu?s de tomarte una bebida con una amiga, ella amablemente paga las bebidas. T? le dices a
ella..."
Below
the
text
of each
scenario
was
an answer
box.
The DCT allows participants to produce P&P in realistic communicative
settings, but the
written nature of the activity also gives them sufficient time to attend to grammatical accuracy
and the opportunity to display knowledge that the cognitive demands of face-to-face
interaction
not
and
Rintell
and
allow
Mitchell
affirm
that
(Bergman
might
Kasper 1993).
(1989)
subjects
responding inwriting can mentally plan what they want to say, thus the final product does not re
flect the trial-and-error planning process that oral speech does. Lastly, the native-speaker corpus
revealed the complete range of P&P functions as established in the literature. Hence, the absence
of certain uses of P&P in the non-native data is not necessarily an artifact of the test itself.
2.3 Procedures
and Taxonomy
One of the challenges in analyzing P&P is the elaboration of a taxonomy of the functions of
the two prepositions. To facilitate comparison between our results and those from previous
studies, we do not deviate significantly from the taxonomy of Guntermann ( 1992) and Lafford and
Ryan ( 1995), which was based on Lunn (1985). One change we made for the classification of para
was that we collapsed the purpose and intended-use categories to one, as there appeared to be
no independent reason to separate the two. This is similar to how Lunn (1985,87) originally posits
this function of para.
1
Table
Taxonomy
Por
for Accurate
Uses
Duration
Examples
por mucho
Motive
gracias
and Para
tiempo, por tres a?os
por, vamos por un caf?
trabajar por m?, $10 por el vaso
por la ma?ana
Exchange/Substitution
In time span
Medium
Formulaic
of Por
expressions
por tel?fono
por supuesto,
por eso_
Para
Purpose/Intended
Beneficiary
(Standard
Para que
use
of) Comparison
comer, dinero para
regalo para mi madre
es dif?cil para m?
para que lleguen a tiempo
comida para
comprar
algo
There are several interesting questions that arise when examining the various uses of P&P,
not the least of which iswhat constitutes/ormw/a/c
expressions. For example, most would agree
an
that porfavor
understanding of the function of the preposition
("please") does not constitute
por by a learner. However, the matter is complicated by some of the other tokens that have been
such as por eso, which present difficulties for
expressions,
previously classified asformulaic
classification. While/?or eso is classified as a lexeme in both of the published studies on P&P, the
use of por in this phrase technically fills the motive function. Consider the following:
1.Por el problema en la oficina, no podr? asistir a lafiesta.
un problema en la oficina. Por eso, no podr? asistir a la?es ta.
2. Hay
of Por and Para
615
Acquisition
In (1), eso refers to the idea of there being a problem in the office, and in each case por expresses
themotive for missing the party. Futhermore, there exists the possibility of para eso in other con
texts, so the learner truly has a choice between/?or andpara. Compare this topor favor, inwhich
para favor is not a reasonable alternative and is not something that learners produce.
In sum, while we maintain a category for fixed expressions, we acknowledge that the defini
In our data totals, we have chosen to
tion of fixed expression requires further examination.
exclude por favor from the percentage of accurate/inaccurate
responses based on the fact that it
does not entail a choice for learners. Given the nature of the DCT, which elicited various speech
acts including requests and complaints, por favor was used frequently by learners, up to 58 times
at Level 2. Therefore, including this highly used lexeme in our analysis would unnecessarily skew
the data.
In addition to the question of the taxonomy of P&P, the categorization of errors raises dif
ficult questions because such analysis requires more than just looking at instances inwhich the
prepositions are either used correctly or confused with one another. There are examples inwhich
is needed (overextensions),
and there are cases in
por or para is used where no preposition
which por or para is used when another preposition or element is required. The converse also
occurs; there are cases where nothing (zero) or another preposition or element
cither por orpara is needed. Lastly, learners can employ ungrammatical syntactic
are difficult to classify as errors involving themisuse of P&P (c.g.,ponelo m?s por
it down"). In these instances, we classify the errors in the category Interlanguage
is used where
structures that
abajo for "turn
(IL). Given all
means
no
these possibilities,
the categorization
is
to facilitate
clear-cut.
However,
system
by
we
use
uses
the
uses
four
Accurate
Accurate
discussion,
groups: 1)
following
involving por; 2)
uses
uses
Inaccurate
Inaccurate
involving para; 3)
involving por; 4)
involving para.
3.0 Data Analysis
3.1 Overall Accuracy
and Native
Speaker Baseline Data
The first issue we want to consider is the overall accuracy of P&P across the four levels, and
then we will look into specifics regarding each preposition, the various uses, and potential areas
of difficulty. In Table 2 below, the right column contains the percentage of accurate uses of P&P
by the four levels of learners. From Level 1 to 4, there is only an increase of 8%, although as one
can see, the increase is not steady and it is actually Level 2 that has the highest percentage of
accuracy (67%). In order to see if the rate of accuracy and inaccuracy was independent of the
level of learners, we ran a Chi-Square test with the data in Table 2. The results showed that there
were no significant differences (%2= 6.2; df = 3; p = 0.101) between the four levels.
The fact that Level 2 outperforms the other two more advanced levels may suggest that the
gains achieved through classroom presentation have somewhat of a short-term effect. The
students in our Level 2 had received instruction on P&P in Spanish 3 and again in Spanish 22,
which may account for their more accurate use of the prepositions. However,
it appears that as
students go on to take classes in composition and eventually in literature and linguistics, the
shift of focus to the macro-level
of meaning from the micro-level
of syntactic forms results in a
decrease in accuracy. Our data, then, seem to show a premature high at the end of the second year
(Level 2), which may represent superficial acquisition, followed by a dip in the third year, similar
to the decrease at the intermediate levels reported by Lafford and Ryan (
1995). Thus, it is possible
that four years of undergraduate study, without a stay in the target culture and with the current
may not elevate students to the level at which
teaching methodologies,
more than a few basic uses of common prepositions.
they are ready to acquire
616
Hispania
89 September 2006
2
Table
Accurate
and
Inaccurate
Instances
Overall:
Por
and
Para1
Accurate
_Level_Accurate_Inaccurate_Percentage
1 (n=108)
2 (n=112)
3 (n=105)
57
75
56
5153%
3767%
49 53%
263
185
4
(n=123)_75_48_61%_
Total (n= 448)
59%
Table 3 compares the number
the total percentages of por (33%)
and Ryan ( 1995) and almost exactly
strates a higher incidence of por.
of tokens and percentages ofpor with those of para. Overall,
and para (67%) coincide exactly with the findings of Lafford
with Guntermann ( 1992). Level 2 is the only level that demon
The similarity across all three studies constitutes a strong
indicator of a degree of stability of por mdpara
in the IL of L2 learners of Spanish, regardless of
whether they have studied abroad or learned exclusively
in a classroom setting.
While these comparisons between groups are interesting, we felt it important to compare the
distribution of P&P with native speakers of Spanish. For this comparison, we used theNS corpus
mentioned
above in order to ensure that the data would be comparable. Interestingly, and
contrary to the learners, the native speakers usedpor (58%) more thanpara (42%). To determine
if there were significant differences or not between the 4 levels and the NS norm, we ran a Chi
Square test to see if each level usedpor with a frequency that was significantly different than the
NS norm of 58%. The difference was shown to be significant for Level 1 (x2= 17.0; df = 1; p =
= 1 =
= 1 =
;p 0.000), and Level 4 (x2= 22.2; df
;p 0.000), while the
0.000), Level 3 (%2=20.6; df
= 1.1 df
was
not significant (%2
difference between Level 2's use (51%) and that of theNS group
;
= 1 =
;p 0.300). Therefore, the learners from Level 2 were the closest to theNS norm in usingpor
more
thanpara,
a finding we discuss below
in Section 3.2.
Table
Total
and
Attempts
_Level_Por_Para_Total
132 (30%)
2
57(51%)
3
28(27%)
_4_33(27%)_90
Por
Percentages:
vs. Para
Por
76(70%)108
55(49%)112
77(73%)105
and Para_
(73%)_123_
448
298(67%)
Total
150(33%)
3.2 Accurate
3
Uses ofpor
striking feature of Table 4 is the apparent accuracy that students display when
as the following discussion will
the totals are somewhat misleading,
However,
por.
using
elaborate. In our study, the most frequent use of por is that of duration. Guntermann (1992) and
Lafford and Ryan (1995) mention that por with duration of time is not accepted by all native
speakers in all contexts. We also obtained the same results from native-speaker judgments. In
fact, one native-speaking
Spanish professor pointed out thatmost textbook explanations of P&P
this
of
usage, when in fact, the majority of native speakers would opt for no
encourage
type
The most
preposition
(i.e. Estuve
tres
semanas
en
Argentina
vs.
Estuve
por
tres
semanas
en
Argentina).
(1992), we considered this usage as accurate, although we will return to
Following
this issue of por plus duration of time in Section 4.0.
It isworthy of note that Level 2 students use almost two times more tokens of por than the
other levels, and twice as many tokens of por for duration of time, while the upper levels begin
incorrectly to employ para. Guntermann ( 1992,183) reported that the intermediate learners in her
study showed a "propensity for usingpor frequently but incorrectly," and our Level 2 students
may be at this stage where they use por with a higher frequency than the other levels, but in this
case, they do so with an accuracy of 79%. Level 2's heavy use of por for duration could be
Guntermann
of Por and Para
617
Acquisition
partially influenced by the text for this level, Al corriente (Blake, Ramos andMarks 1998), which
?as most texts do? includes "time durations" as one of the uses of por along with examples and
exercises.
The motive function of por is the second-most-frequent
category. Due to the fact that the
contained expressions of gratitude, there are a number of tokens of gracias por ("thanks
for"), which comprise the vast majority of examples in this category. As a group, however, it does
not appear that students are actually acquiring this usage given that even at the advanced level,
gracias para appears with more or less the same frequency as gracias por.
The accurate use of por for substitution and exchange by Level 1 is in need of some explana
DCT
tion. The scenario from the DCT that elicited these tokens required the participants to ask their
roommate to clean the apartment for them, even though it is their turn to do so. In Spanish the text
scenario includes por ti. Thus, in theory the learners have this correct usage as a basis tomodel
their response. Nevertheless,
only two Level 1 learners and one Level 2 learner uscdpor correctly
in this sense of substitution: all the other uses in the data showpara ti. Itwould be difficult to de
termine whether these three particular learners have acquired this use of por, but the distinction
between doing something por mi/ti (in place of) or para mi/ti (for the benefit of) is one of the
subtleties with which learners struggle since they both collapse to "for" inEnglish. The fact that
all the students from Levels 3 and 4 opted for para mi/ti, even when the scenario included por in
the text, demonstrates the extreme difficulty in acquiring this usage.
Table
Accurate
4
Uses
of Por
_Category_Level
1_Level
239 11
11
Duration
Motive
2
Exchange/Substitution
Medium
2_Level
11
3_Level
9
4
14
3 10 0
10
10
Time
span 0 10 0
por eso
3 4 3 1
por supuesto
por
accidente_1_0_0_0_
Total
accurate_21_45_24_26
Total
overall_32_57_28_33
Percentage
0
5
0
2
65.6
79
8679
Por is considered to be themore opaque preposition of the pair (see Guntermann 1992; Lunn
1985), and the fact that the accuracy percentages of por are higher than those ofpara is rather
In addition to the issues just discussed, expressions such as por eso and por supues
misleading.
to are likely memorized by learners in chunks and may not demonstrate true
acquisition of the
meaning of por. Furthermore, the para for por errors discussed below are the most frequent
misuses ofpara and ultimately signal the learners' inability to grasp all the functions of por.
3.3 Accurate
Uses of para
As was the case in the two abroad studies, the purpose category for para showed the most
consistent use among all students, accounting for 81% of the correct uses across all four levels
(see Table 5). This number increased with level, from 77.8% at Level 1 to 89.8% at Level 4.
Although one might expect more varied uses of para at the higher levels, the DCT likely con
tributed to the inflated use of purpose since this function coincided with the
particular discourse
tasks. However, theNS use of para for purpose accounts for only 72% of thepara tokens, lower
than all the levels, especially Level 4 (89.8%). One possible explanation for this may be that
learners use this function because they come to associate para with the infinitive
(e.g. para
trabajar), favoring this syntactic structure over others. For example, para is often inter
changeable with a in sentences such as the following:
618
H?spanla 89 September 2006
3. Voy a la biblioteca para estudiar. "I'm going to the library in order to study."
4. Voy a la biblioteca a estudiar. "I'm going to the library to study."
There are many examples of this type in the data. At Level 4, roughly 30% of the tokens ofpara
for purpose could be substituted with a or some other element.
The beneficiary category is the only other category that shows any consistent usage among
the four groups. Ifwe combine the purpose and beneficiary categories, they account for 94.6% of
all accurate uses ofpara. Ifwe compare Level
variety of uses of para. At Level 1 beneficiary
1with Level 4, we see very little difference in the
and purpose account for 97.2% of accurate uses,
and at Level 4 the number is only slightly lower at 95.8%.
These data indicate that the beneficiary and purpose uses of para are acquired very early in
the learning process, and we see relatively little change throughout a four-year sequence of uni
versity coursework. There is only a slight increase in overall accuracy from Level 1 to Level 4,
thus indicating that students are not correctly employing other uses of para beyond those
acquired at the beginning
levels.
Table
Accurate
Uses
5
of Para
1_Level
Category_Level
Use
Purpose/Intended
7 3 6 3
Beneficiary
2824 44
(Standard of) Comparison
Para
que_0_2_0_2
Total accurate_36_30_32_49
0112
Total
overall_76_55_77
4
90
47.3
Percentage
3.4
3_Level
2_Level
54.5
41.5
54.4
Inaccurate Uses ofpor
The most frequent inaccurate use of por (Table 6) is that of "por for zero." All the instances
of "por for zero" in our data involve learners using Verb +por, apparently a transfer phenomenon
from the parallel structures Verb + "for" in English, while Spanish uses zero. The errors that
appear are with the verbs pagar, sentir, pedir, and esperar. These verbs take a direct object in
Spanish {pagar las bebidas, sentir llegar tarde, pedir una copa, esperar el autob?s), but a pre
positional phrase in English with "for" (pay for the drinks, sorry for arriving late, ask for [or
"order"] a drink, wait for the bus). Due to the fact that learners are transferring their English
to Spanish, these tokens do not necessarily reflect problems with the semantics of
knowledge
as the data show, this is a problematic issue for students; in fact, Padilla-Falto
Nevertheless,
por.
most of the errors she found in her oral data entail problems of transfer from
that
mentions
(1996)
English to Spanish.
"
are not as frequent in our data as cases of "para for por" Given
Instances of "por for para
that Level 1 shows the highest frequency of this error (36%), this pattern may be part of the phe
nomenon mentioned
inGuntermann (1992) inwhich "at the earliest stages of learning there is a
to
equate por with English for" ( 184). However, since Level 1 students were nearing the
tendency
end of their first year of study, they are probably moving beyond this stage of initial contact.
of Por
Acquisition
Table
Por:
por
for
por
for zero
por
for de
por
por
Inaccurate
2_Level
0
4
3_Level
1
3 10 3 5
1 0 0 0
4 10 0
for para
for en 1 0 0 0
JL_2_1_1_1
inaccurate /
IL_H_12_4_7
Total
33
Totaler
32_57_28
34.4%
Percentage
3.5
619
IL Usage
and
1_Level
_Category_Level
0
0
a
and Para
6
Inaccurate
21.0%
14.3%
21.2%
Uses of para
all levels, the students in this study show an error rate of 50.7% withpara
(Table 7),
more than twice the error rate with por (22.7%). Although 40% of the para errors involved the
substitution ofpara for por, a closer examination of the data will reveal that confusion between
por and para does not seem to be a significant issue for learners, as por for para occurred only
Across
five times in the study, and only once after Level 1. Thus, the students are overextending
the use
of para to contexts in which por is the correct preposition, but beyond the beginning stages,
they do not show a tendency to use por for para. Further evidence for an overextension of para
among our participants is the 28.5% ofpara errors that came from the use of para when the
context called for zero, in addition to the 14.5% of para errors stemming from the use of para in
of another preposition or element. In other words, the learners in our study show a
significant overextension of para tomany other uses at all levels.
As pointed out in section 3.4, themajority of errors with por came from por for zero, which
may represent nothing more than a problem with translation (the na?ve lexical hypothesis, Gunter
mann 1992), inwhich students assume that for every word in a sentence in their native language,
place
there must be one word
in the L2. Beyond this particular error with por, the subjects used por
This
suggests that para carries a heavy semantic load for L2 learners, but as
quite accurately.
to
understand
various uses of por, these uses emerge with a high degree of accuracy.
they begin
Table
Para:
Inaccurate
7
and
IL Usage
_Category_Level
1_Level
para
para
para
para
para
3_Level
4
11 12 18 20
for de 2 0 4 1
for que
114 5
fora 0 0 10
for duration
4 111
for zero 13 5 13 12
para
zero for para
a for para
5 2 2 0
1110
IL
2
que for
para_1_0_0_0_
Total inaccurate /
IL_40_25_45_41
TotdX
para_76_55_77_90
Error percentage
4.0
2_Level
for por
Implications
52.6%
3
45.5%
12
58.4%
45.5%
for Teaching
There is a tradition in the organization
vs. estar, preterit vs. imperfect, subjunctive
P&P, such a contrast does not reflect the
Guntermann 1992). In order to present two
of Spanish grammar to present items in contrast {ser
vs. indicative, saber vs. conocer), but in the case of
reality of how they occur in realistic discourse (see
entities contrastively, one must assume that if it's not
620
H?spanla 89 September 2006
one, it's the other. Otherwise, there is relatively little reason to contrast the two elements. For all
of the above-mentioned
grammar points, with the exception of P&P, such a rule generally holds.
ifwe examine the results of our study, we find that of the 34 errors with por, only five
However,
should have been para. In other words, if students found por marked wrong on their paper and
simply changed it topara, they would still be wrong more than 85 percent of the time. In fact, they
would be much better off simply deleting por, which would result in the correct structure nearly
two-thirds of the time.
As we noted previously, the incorporation of P&P in first- and second-year texts is unique
among Spanish prepositions in that they are presented in opposition and with up to 14 individual
functions, often to be covered in one class period. In contrast, all other prepositions are pre
sented independently, embedded in the context of other communicative or grammatical foci. For
example, the function of de to express possession or origin is often incorporated into the presen
tation of the verb ser so that students can form basic sentences such as El libro es de Juan and
Yo soy de Venezuela. These are not the only functions of de, but rather they are themost common
functions relevant to the verb ser. Other uses of de, such as la chica de la camisa azul, llenar el
coche de gasolina, or an expression like de ni?o are not commonly covered explicitly in be
ginning texts.
Why, then, must we bombard students with more than a dozen uses of P&P, often in one
class session, beginning in Spanish 1? If the justification is that "for" inEnglish always translates
to por ox para, this is naive since zero and other prepositions are possible. In addition, the re
verse also is not true; that is, P&P do not always translate to "for" in English (e.g.,por semana
"per week" andpara comprar "in order to buy"). Thus, in order to avoid the problem of cognitive
overload thatmost likely accompanies the presentation of P&P, we recommend a sequential ap
proach, distributing relevant functions at various points throughout the curriculum rather than
into one brief grammar lesson. By presenting the
forcing all uses of the two prepositions
individual functions of P&P, the students receive richer input that would allow them to focus
their attention and attend to a particular function, a factor that plays a crucial role in the acquisi
tion process (Doughty andWilliams
1998, Schmidt 1993).
Under the current organization, students are not afforded the opportunity to acquire one use
before moving on to another. Instead, students are forced to seek practical solutions to prepare
for exams. For example, even some discussion aimed at teachers recommends that students
should "learn the rules for one of the prepositions completely" and "when one of those rules
doesn't apply, they should use the other preposition" (Mason 1992, 198). While this approach
works for students preparing for an exam on which they must choose between P&P, it ignores
several of the important facts that we have been discussing. Furthermore, these practical strate
gies do not necessarily aid in the long-term acquisition process, as the data from our study and
others indicate (Padilla-Falto 1996, Yee-Wun Lam 2003).
As mentioned
in Section 3.2,94.6% of the accurate uses ofpara were accounted for by the
purpose and beneficiary categories. A similar situation applies to por, where the categories of
"duration of time," "motive" and "formulaic expressions" comprise 94% of the accurate uses.
This limited range of P&P functions contrasts sharply with the NS corpus collected using the
same instrument, which contains uses that are absent in the learner data. For example, regarding
para, native speakers made reference to temporary goals or deadlines (para otro d?a, para
spatial meaning (voy para otro rumbo), and certain specialized uses (no
estoy hoy para X). Regarding por, there are NS uses involving spatial meanings
(?Por d?nde
andas? P?sate por mi casa, ?Qu? andas haciendo por aqu??) and more expressions (por si, por
lo que, por lo cual, por completo). By redistributing the presentation of P&P, the students would
be able to attend to the wide range of native-like uses, receiving many examples of each use. We
also need to consider that if students learned the textbook rules for P&P perfectly, they still
would commit many of the most common errors such as busco por mi libro, es importante para
ma?ana, para el jueves),
estudiar mucho, and awkward uses of por with time duration thatwould
an hace + time expression + que.
likely be rendered with
Acquisition
of Por
and Para
621
This leads us to another strategy thatmay be effective for overcoming some of the difficul
ties in learning and teaching P&P; that of incorporating more productive formulas into textbook
there is some debate in the literature regarding the extent to which
presentations. Although
formulas contribute to language acquisition, there are studies that highlight the importance of
routines or conventionalized
language forms as part of the learning process (Raupach 1984,
Schmidt 1993, Yorio 1980). In our data, some evidence exists that learners are not acquiring set
formulas that contain eitherpor oxpara. Gracias por is one example, where even at Level 4,50%
of the learners continue to use gracias para. One would expect gracias por X to be a fairly regular
occurrence in the input learners are exposed to, even in classroom discourse. However, the fact
that fourth-year students still perform at an accuracy rate of 50% (equal to a 50% probability of
guessing gracias por/para correctly) suggests thatmore repeated and explicit attention to form
function may be required. Padilla-Falto ( 1996), for example, also claims that "there is evidence in
favor of the effectiveness of the grammar consciousness-raising
tasks" (193).
The use of formulas would also address instances inwhich learners rely on transferring syn
tactic structures from their LI. A recurring example from our data involves the non-target usage
ofpor with time duration, when the formula hace + time expression + que would be the preferred
this structure does not use por or para in Spanish, English
by native speakers. Although
learners
speaking
repeatedly employ versions of no te he visto por muchos a?os. Although we
did not consider this an inaccurate usage ofpor in our analysis, learners are clearly not acquiring
the target structure for expressing "having or not having done something for X amount of time."
This
involves P&P because if the justification for teaching these two prepositions together is
based on the fact that they are often equivalent to "for" inEnglish, we also need to teach learners
the other common ways that "for" may be expressed in Spanish. One of these ways involves the
use of hace que with time expressions. In other words, teaching P&P in
opposition and at the
of
other
assume
forms
to
mislead
students
that
"for" is equal to no
expense
may
prematurely
other than eitherpor oxpara. Similarly, there is the problem inwhich other time expressions, often
involving "for" inEnglish, may be preferably expressed with zero in Spanish, an issue that needs
to be further investigated.
Given that por is the most problematic and opaque preposition of the pair, and taking into
consideration that students overuse para, an effective teaching strategy would be to focus more
on specific uses ofpor at the lower levels, exposing students to each of its distinct functions at
various points in the first and second-year curricula. For example, with the exception of gracias
por, students seldom use por to express motive. This function would be highly useful for basic
communication
(llegu? tarde por la lluvia, suspendi? el examen por no estudiar), and a lesson
incorporating multiple
examples of this particular use would be highly productive, giving
students the opportunity to acquire one function before moving on to another.
5.0 Conclusions
Our study indicates that in a classroom setting, learners show only an eight percent im
provement in accuracy with P&P over the course of a four-year university program, despite
explicit grammar explanations and continued exposure to the target language. The fourth-year
participants in our study, with an accuracy rate of 61%, do not approach the level of 94%
achieved by the superior learners in Guntermann (1992) or the 82% reached by the advanced
learners inLafford and Ryan ( 1995).
In addition, the accurate uses of P&P by classroom
that remains relatively stable over the four years. More
learners lie within a very limited range
than 90 percent of the accurate uses of
para fall under the categories of beneficiary and purpose, while more than 90 percent of the
accurate uses ofpor can be accounted for by the categories of motive, duration of time, and fixed
of para to many different
expressions. The two most notable problems are the overextension
environments and the inaccurate use of Verb +por due to transfer from English. Most modern L2
textbooks do not address either of these issues. They continue to present P&P in opposition with
622
H?spanla 89 September 2006
up to seven functions for each preposition. A sequential approach to the presentation of the
various functions of P&P would fit well with current research on acquisition. Nevertheless, much
more research needs to be done to ensure the effectiveness of any modifications. A crucial issue
is to determine the appropriate point at which learners acquire each of the various functions of a
given preposition. While some uses of prepositions may be acquired very early, other uses may
be acquired much later. In order to investigate
be carried out on each individual preposition.
these details,
longitudinal
need to
studies would
Acknowledgments
com
We
reviewers
for their insightful
and the anonymous
offer sincere thanks to Dr. Joseph Collentine
as well as to Professors
and Germ?n
Robert Raymond
assistance. We assume
ments,
Pliego for their statistical
for any errors that might
sole responsibility
appear.
NOIE
on this table and determining
of our manuscript
recommended
the
running an ANOVA
would give us essentially
the
each group. The statisticians we consulted verified
that an ANOVA
same p-value
test. Furthermore,
that looking for variance
that we have with the Chi-Square
"with
they claimed
in this table are
of a response from the mean.
Since our responses
in a group" measures
the typical distance
as "accurate/inaccurate,"
of an individual
their mean
is just the fraction correct,
thus the distance
recorded
is not a useful
concept.
response from that mean
*One reviewer
within
variance
WORKS CITED
Bergman,
M.
L.,
and
Gabriele
Apology."
Interlanguage
UP.
82-107.
Robert
Blake,
Colombi,
Maria
J., Alicia
Cecilia,
Mifflin.
Houghton
Catherine,
Doughty,
Kasper.
Pragmatics.
and Martha
Ramos,
Jill L. Pellettieri,
and Jessica Williams.
Cambridge
Cambridge:
Gail.
Guntermann,
(1992).
(1993).
"Perception
Eds. Gabrielle Kasper
A. Marks.
and Maria
(1998).
Focus
and
Performance
and Shoshana
(1998). Al corriente.
Isabel Rodriguez.
on Form
in Native
Blum-Kulka.
and Nonnative
New
Boston:
McGraw-Hill.
(2001).
Palabra
in Classroom
Second
York:
abierta.
Language
Oxford
Boston:
Acquisition.
UP.
over Time:
Part 1, por and para."
"An Analysis
of Interlanguage
Development
177-87.
75.1:
Hispania
Eds. Susan Gass
Eric. (1983). "If at First You Do Succeed..."
Kellerman,
Acquisition.
Input in Second Language
House.
345-53.
and Carolyn Madden.
Rowley, MA: Newbury
in a Study-Abroad
"The Acquisition
of Lexical Meaning
Barbara A., and John M. Ryan.
Lafford,
(1995).
78.3: 528-47.
The Spanish
Context:
por and para." Hispania
Prepositions
Lunn,
Patricia.
Mason,
Keith.
(1985).
"The
Hispania
I. (1996).
Padilla-Falto,
Olga
the Spanish Prepositions
Manfred.
(1984).
Raupach,
of por
Mnemonics
Semantics
"Successful
(1992).
197-99.
75.1:
Indiana University.
and para." Diss.
Commands
for por ?para and Affirmative
with
Pronouns."
on the Acquisition
and Field Dependence
of
of Formal
Instruction
Diss. Georgetown
por and para"
University.
Production."
Second
in Second
"Formulae
Language
Language
Speech
114-37.
G. Narr.
and Manfred
Dorothea Mohle,
Productions.
Eds. Hans W. Dechert,
Raupach.
Tubingen:
An Inquiry into Method."
and Apologies:
J. Mitchell.
Ellen M. and Candace
Rintell,
(1989). "Studying Requests
and
Juliane House
Eds.
Shoshana
and Apologies.
Cross-Cultural
Blum-Kulka,
Requests
Pragmatics:
248-72.
Gabriele Kasper. New
Publishing.
Jersey: Ablex
"The Effects
and
"Consciousness,
Interlanguage
Pragmatics."
Interlanguage
Learning
(1993).
New York: Oxford UP. 2X-A2.
and Shoshana Blum-Kulka.
Gabrielle
Kasper
Boston:
Mu?oz.
and Elias Miguel
Jeanne Egasse,
Andrade,
(1998). Dos mundos.
Terrell, Tracy D., Magdalena
McGraw-Hill.
for
of Interrelating
Rules
The Effectiveness
Yee-Wunn
Lam, Yvonne.
"Challenging
(2003).
Prepositions:
Schmidt,
Richard.
Pragmatics.
Yorio,
Eds.
in Spanish as a Second Language."
Teaching por and para
Forms
"Conventionalized
Carlos.
Language
(1980).
14.4: 433-42.
TESOL Quarterly
Competence."
Diss.
and
of Toronto.
University
of
the Development
Communicative