The Modern Language Journal

(1993). 77.1: 45-57

InputProcessingand Second
LanguageAcquisition:A Role
forInstruction
BILL VANPATTEN and TERESA CADIERNO
Departmentof Spanish, RomanskInstitut
Italian & Portuguese AarhusUniversitet
Universityof Illinois NielsJuelsGade 84
Urbana, IL 61801 8200 AarhusN
E-Mail: vanpatte@uxl .cso.uiuc.edu Denmark

AS ATTENTION TURNED TOWARD THE Chomskyantheoryas a means of pinpointinga

role of input in second language acquisition role for explicit instruction(e.g., 16; 41; 42).
(SLA) to address questions surroundingacqui- Since the early seventies, when research first
sitionorders and transitionalstages of compe- began to look at SLA, evidence has slowlyac-
tence (e.g., 14), some researchers began to crued that explicit instruction can facilitate
questionthe role of explicitinstructionin gram- SLA.
mar (e.g., 20; 21; but see also 27 for earlier The purpose of the present paper is not to
discussion).' For these researchers,explicitin- enter into the debate on whetheror not there
structionin grammar (i.e., deliberate attempts should be a role forexplicitinstructionin gram-
to focuslearners'attentionon the structureand mar in language teaching. Following Garrett,
forms of language coupled with practice in we adopt the position that research on explicit
manipulatingthose forms and structures)was instructionshould consider what kind of gram-
thoughtto aid verylittlein the acquisitionpro- mar instructionlanguage learners can use. It
cess. SLA itselfwas viewed as the process of may verywell be that previous research is lim-
extractingdata fromthe input duringexposure ited because both the grammar that has been
to the speech of another person. Classroom taught and the manner in which it has been
learning was shown to resemble, if not be the taughtdo littleto affecttheprocessesthatunder-
same as, non-classroomlearningas far as route lie acquisition.In the presentpaper, we recon-
of acquisitionwas concerned (e.g., 6; 9), and to sider the nature of explicitinstructionvis-a-vis
a certain degree, the acquisition of a second the processes involved in language acquisition
language also resembled firstlanguage acquisi- and language use. Specifically,we willfocus on
tion (e.g., 43). the relationship between explicit instruction
Partlyas a reaction to Krashen and partlyas and input processing.We report the resultsof
naturalgrowthin the field,researchand discus- a studyon the effectsof two differenttypesof
sion began to suggest that explicit instruction instructionon the developing knowledge sys-
was beneficial(e.g., 24; 30), thatitwas necessary temof the L2 learner: instructionas the manip-
(e.g., 15), that exposure to input was not ulation of output and instructionas structured
enough (e.g., 34), and that classroom learning or focused input processing.
(regardless of focus of instruction) resulted
in "more acquisition" in learners than non- SLA AS INPUT PROCESSING
classroom environments (e.g., 28). Research
also suggestedthatthe effectof explicitinstruc- SLA is multifacetedby nature, and more
tion was constrained by learnability(e.g., 31), than likelyno single theoryof language or psy-
while other research began to incorporate chology will be able to capture how language
acquisition happens (5; 41). Whatever one's
theoreticalbent, most would agree that SLA
TheModernLanguageJournal,77, i (1993) consistsof sets of processes as sketchedin Fig-
0026-7902/93/45-57 $1.50/0
ure I. What Figure I attemptsto capture are
? 1993 TheModernLanguageJournal
three distinguishablesets of processes in acqui-
46 TheModernLanguageJournal77 (1993)

FIGURE I FIGURE II
Processesin SecondLanguageAcquisition TraditionalExplicitGrammarInstruction
in
ForeignLanguageTeaching
I II III
input -- intake -- developing system -- output input - intake - output
developing system -

t
focused practice
sition. The first(I) converts input to intake.
From intake the learner must stilldevelop an
acquired system;thatis, not all of intakeis auto-
maticallyfed intothe acquired system.The sec-
ond set of processes (II) then includes those FIGURE III
thatpromotethe accommodationof intakeand Processing inForeignLanguageTeaching
Instruction
the restructuringof the developing linguistic
system(25; 41; 42). Finally,researchon output intake -- developing system -

t
input --- output
reveals thatlearner language is not a directre-
flectionof acquired competence. Thus, a third
processing mechanisms
set of processes (III) mustbe posited to account
forcertainaspectsof language production,e.g.,
monitoring,accessing,control,and so on. focused practice
Input processing is concerned with the first
set of processes; thatis, the conversionof input
to intake. While differentperspectivescan be
taken on input processing (see 13; 35; 36), we quisition must contain meaning to which the
will use the notion of "form-meaningconnec- learner attends for its propositionalcontent.So
tion" to discuss the processes involved in the called negativeevidence and much of corrective
conversion of input to intake. That is, input feedback focused on grammaris not language
processing,as used here, involvesthose strate- that contains propositional information.They
gies and mechanisms that promote form- do not enterinto the definitionof input thatwe
meaning connections during comprehension use here. We do not view input as explanations
(see 36 for a discussion of input processing). about how language works. In short,input for
Comprehension and input processing are not us containsreferentialmeaning.
equivalent terms. Given that input processing In order to understand the connection be-
involves making form-meaning connections tween input and instructionthatwe would like
that can be used for acquisition (note the im- to make, we must remember how traditional
plicitrelationshipbetweenI and II in Figure I), explicit instructionin grammar occurs. Nor-
sufficientargumentsexist that comprehension mallythisinstructionfocuses on the manipula-
does not necessarilylead to acquisition(e.g., 8; tion of learner output. That is, instructionoc-
35; 36; 42). curs by explaining a grammaticalconcept and
At thisjuncture we mustdefine"input,"since thenhavinglearnerspracticeproducinga given
the term has been used in a varietyof ways. structureor form (see Figure II). Given the
Krashen (20; 21, and elsewhere) has suggested ratherimportantrole that input plays in SLA,
thatcomprehensible input is a necessaryingredi- the value of grammaticalinstructionas output
ent for acquisition.White (42) has argued that practice is questionable if the attemptof the
incomprehensibleinput may trigger the ac- instructionis to alter the nature of the devel-
quisition of some aspects of the grammar. oping system.Note thatin Figures I and II, the
Schwartzsuggests that primarylinguisticdata inputdata (i.e., intake) flowintothe developing
defined as utterances in the target language system.In other words,the arrowsgo fromleft
constitutethe input necessaryfor the language to right,not fromrightto left.Ratherthan ma-
module described by Fodor. What is common nipulatelearners'output to effectchange in the
to all these conceptualizations-and what de- developing system,instructionmight seek to
fines input as we use it in this paper-is that change the waythatinput is perceivedand pro-
input must be language that encodesmeaning. cessed by the learner. This approach to instruc-
That is, the input necessary for language ac- tion is depicted in Figure III. Theoretically,
Bill VanPattenand TeresaCadierno 47

altering input processing should have a sig- The result in the learner's developing system
nificantimpact on changing the internalized as revealed by output is an absence of object
knowledge. pronouns or the misuse of object pronouns as
To research such a hypothesisone mustfirst subjects of a sentence, incorrectplacement of
identifythe strategies and mechanisms used object pronouns, the (re)settingof an incorrect
during input processing. While much is not parameter(thatis, the learner may assume that
knownabout input processing,some discussion Spanish is [-null-subject]),the absence of the
appears in the literature about how learners case markera, and difficulty in the acquisition
process input (e.g., 4; 8; 13; 17; 18; 32; 37). In of a certainclass of verbsthatobligatorilyplace
addition, research in child L1 acquisition of- subjectsin postverbalposition (e.g., gustar).In
fers discussion of possible strategiesused by short,learners fail to see that Spanish is not a
L2 learners while processing input (e.g., 29). rigid SVO language.
One strategythat has received considerable
attention involves the use of word order to
PRESENT STUDY
assign argument structureto an input string.
Evidence from child L1, and both child and The present studyexamines the outcome of
adult L2 studies, shows that early and inter- explicitinstructionin processinginput. That is,
mediate learners assign agent statusto the first unlike traditionalFL instruction,we seek to al-
noun(phrase) of a stringand object statusto the ter the strategyby which learners make form-
second noun(phrase) (2; 3; 7; 12; 23; 26; 38). meaning connections when exposed to input
While this may be a useful L2 strategywhen strings(compare Figures II and III). In the
applied to English input sentences (except for studyreported here, we discuss the impact of
passives and clefts),it is not such a useful strat- attemptingto alterlearners'processingof input
egy for Spanish. Spanish has flexibleword or- containingnon-SVO order. As partof the study
der allowinga numberof surfacestructurepos- we compared three groups. The firstreceived
sibilities:SVO, SOV, OVS, OV. Given that traditional explicit instructionin object pro-
ambiguity can result, Spanish uses the case nouns. The second received "processing" in-
markera to identifyobjects when both are ca- structionon the same. The third received no
pable of performingthe action: El senior siguea explicitinstructionat all regardingobject pro-
la sefiorala sigueel hombre,
la seaiora./A "The man nouns. We asked the followingquestions at the
followsthe woman." outsetof the study: 1) Does alteringthe way in
Withcliticobject pronouns, word order is less which learners process input impact on their
flexiblewith obligatorypreverbal position for developing systems?2) If there is an effect,is it
object pronouns when the verb is a simple finite limitedsolelyto processingmore input or does
verb.The subjectmay be placed beforeor after instructionin input processingalso have an ef-
the verb depending on features of discourse, fecton output? 3) If there is an effect,is it the
style,and pragmatics.The followingtwo sen- same one that traditional instructionhas (as-
tences are both translatedas "The man follows suming an effectfor the latter)?
her." Subjects.Three second year universitylevel
Spanish classes at the Universityof Illinoiswere
El sefior la sigue. selected at random fromthose available in the
The man-SUBJ follows.
her-OBJ Spring of 1991 for inclusion in the present
La sigue el sefior. study.Each class was randomlyassigned to one
follows the of the treatmentgroups. Group N (no instruc-
Her-OBJ man-SUBJ.
tion) consisted of eighteen subjects, Group P
Research on learners of Spanish has shown (processing instruction)consisted of nineteen
that input strings in which subject-object subjects,and Group T (traditionalinstruction)
(agent-object)order is reversedare misassigned consistedof eighteen subjects.Pretestingelimi-
argument structure(e.g., 22; 23; 38). That is, nated several subjectsin each group, as did re-
preverbal clitic object pronouns are misinter- current absenteeism from post-testingphases
preted as subjects and postverbal subjects are (see below), so that in the final analysis each
misinterpretedas objects as in the following group had seventeen, seventeen, and fifteen
example: subjectsrespectively.
Subjects were enrolled in a program with a
La sigue el seior. communicativemethodology modeled on the
*She follows the man. Natural Approach where emphasis is placed on
48 TheModernLanguageJournal77 (1993)

developing communicativeskillsin the second sistedof a fullnoun and when it consistedof a

language. Classes met fourdays a week withthe clitic object pronoun; and 2) having the stu-
bulk of class time spent on interaction,lis- dents respond to the informationalcontentof
tening,and reading. The grammarinstruction OV strings.The instructionalpacket thatcom-
of the regular curriculumwas limited to text- prised thistypeof instructionwas based on the
book and workbook exercises done as home- manual that accompanies ?Sabias que. . .?: Vol-
work.However,throughoutthe timeperiod for umeI (39).
this experimentation,subjects were not sched- In processinginstruction,the presentationof
uled to receiveexplicitinstructionin objectpro- the direct object pronouns firstcontrastedthe
nouns and word order. grammaticalconceptsof object and subjectof a
Instructional
Packets.Since the aim of the pres- verb,and then presented both subject and ob-
ent investigationwas to examine the impact of ject pronouns. For example:
differenttypes of instructionon the learners'
SUBJECT OBJECT
developing system,two instructionalpackets me
were constructedfor use during the instruc- yo
Yo comprendoa mi Mi hermanomecom-
tional treatment.These packets reflectedtwo
differentapproaches to the teaching of the hermano. prende.
cliticdirectobjectpronouns. The firstapproach (I understand my (My brotherunder-
brother) stands me)
consistedof traditionalgrammar teachingand
oral practice, while the second, called pro- The presentationof the pronouns was fol-
cessing instruction,involved teaching the sub- lowed by explanations of importantpoints to
jects to processinputsentencesdifferently from keep in mind about the pronoun position in
the strategiesdescribed earlier. Spanish. Studentslearned that in Spanish, un-
Traditional instructioninvolved presenting like English, it is possible to have such se-
the subjects with explanations concerning the quences as object pronoun-verb-subject.
form and position of direct object pronouns
Lo llama Maria.
withinthe sentenceand then givingthem prac-
tice in how to make sentences withthose pro- him-OBJ calls Mary-SUBJ.
nouns. Presentation involved a paradigmatic "Mary calls him."
chart that included all persons (both singular In the sectionon the object markera, subjects
and plural) and an explanation of what objects in the processinggroup learned'thatfullobject
and object pronouns are. The presentationalso nouns may be placed before the verb:
included a descriptionwithexamples of correct
A Maria la llama Juan.
placement of object pronouns within a sen-
tence. Subjects were taught that object pro- Mary-OBJ her-OBJpro calls John-SUBJ.
nouns alwayspreceded simpleconjugated verbs "Johncalls Mary."
but could be placed at the ends of infinitives Two types of activitiesfollowed the presenta-
and presentparticiples. tion and explanations of the object pronouns.
In terms of practice, traditionalinstruction One typehad subjectslisteningto utterancesor
involved moving the subjects from mechanical reading sentences and then somehow demon-
formoriented practice (oral and writtentrans- stratingthat theyhad correctlyassigned argu-
formationand substitutiondrills) to meaning- ment structure to the string. This typically
ful practice(oral and writtenquestions, simple involved selectingthe drawing that best repre-
sentence formation)and finallyto more open sented what was heard/read or selecting the
ended communicativepractice (oral and writ- best Englishrenderingof the Spanish sentence.
ten question and answer, conversation).At all The second type of activityhad subjects re-
times the traditional instructionfocused the spond to the content of an utterance or sen-
learners on producingthe targeted items. The tence by checking"agree" or "disagree," "true
instructionalpacket thatcomprised thistypeof for me" or "not true for me," and so on. In
instructionwas based on Puntosde Partida (19) several activities subjects read a very short
and on itsworkbook(1). Examples of the types passage in which subsequent to reading the
of activitiespresent in traditional instruction passage, sentences with object pronouns were
are presented in Appendix A. highlightedand subjectswere asked what those
Processinginstruction,on the other hand, in- particularutterancesmeant. We should stress
volved: 1) teachingthe subjectshow to correctly here thatat no pointdid processinginstruction
in-
interpretOVS strings,both when the O con- volve theproductionof thepronounformsby the
Bill VanPattenand TeresaCadierno 49
TABLE I Lo saluda la chica.
Summary of Traditionalvs. Processing
Instruction him-OBJ greets the girl-SUBJ
"The girl greets him."
Traditional Processing
Paradigmatic Non-paradigmatic
Focuson output Focuson input For the interpretationtasks, subjects were
Somefocuson meaning Meaningalwaysin focus asked to match each sentence theyheard with
one of two pictures that were simultaneously
learners.In addition, the presentation of the presented on an overhead projector. The two
pictures represented the same action, the dif-
pronouns and the activitiesthat followedthem ferencebetweenthem being who the agent was
were broken up into two sectionswithfirstand
second person singular and firstperson plural and who the object of the verb was. For exam-
ple, for the sentence Lo saluda la chica,subjects
constitutingthe firstsection and third person had to choose betweena pictureof a boywaving
singular and plural constitutingthe second to a girl (who did not wave back) and a picture
section.
of a girl waving to a boy (who did not wave
Vocabulary, which consisted of highly fre-
back).
quent items, was the same for both types of The production task was based on activities
instruction.The amount of practicethatlearn-
ers received (i.e., the number of timesone pro- used in traditionalinstructionand included five
duced a sentence in the traditionalgroup vs. items,each of whichconsistedof an incomplete
the number of times one interpreted or re- sentence.The subjects'taskwas to completethe
sentence according to a visual clue. The test
sponded to a sentence in the processinggroup) items were constructed along the following
was also roughlythe same forboth instructional
treatments. Differences between traditional lines:
and processing instructionare summarized in El chico piensa en la chica y entonces
Table I, and specificexamples of activitiesfor
processinginstructionare presentedin Appen- "The boy is thinking about the girl and
dix B. then
Pre- and Post-Tests.A pretest/post-test proce-
dure using a split-blockdesign (see below) was Each item was accompanied by two drawings
used as the means of assessing the effectof in- that depicted the content of the sentence. In
struction.All tests (i.e., the pretest and the the above example, the pictureswere one of a
three post-tests)consisted of both interpreta- boy sittingat home thinkingabout a girl and
tion tasksand writtenproduction tasks.Since a the second showed the boy calling the girl on
sole focus on interpretationwould have biased the phone. It was assumed thatthe second pic-
resultsin favor of those subjects in processing ture would guide the subject to create a sen-
instruction,a production task would either be tence in which an object pronoun should be
neutral or favor the production group, thus used, as in the above example: "The boy is
counterbalancingthe other task. Between the thinkingabout the girl and then he calls her."
interpretationand production tasks, a dis- While the visual cues forthe interpretationtask
tractor task consisting of writinganswers to were projectedfroman overhead projector,the
questions unrelated to and not containingthe drawingsused for the productiontask were on
grammaticalitem under studywas given to the the page along witheach item.All testitemsfor
subjects. both task types involved simple present tense
The interpretationtasksforall testsconsisted sentences.
of fifteenaural sentences. Five of these sen- Instructional
and Data CollectionProcedures.A
tences consisted of SVO word order sentences split-blockdesign was used that included four
which served as distractors.Of the remaining versionsof the same testdescribed previously:
ten testsentences,fiveconsistedof sentencesof A, B, C and D. Half of the subjects received
the followingtype: versionA as the pretestand half received ver-
Al chico lo saluda la chica. sion B. To ensure that subjectsunderstood the
The boy-OBJ him-OBJpro greets the girl nature of the interpretationtask, they were
"The girl greetsthe boy." givena practiceitem.To controlfor familiarity
of vocabulary,a listof Spanish-Englishequiva-
The other fivesentences were of the following lents was provided to subjects prior to testing.
type: Subjects were given two minutes to study the
50 The ModernLanguageJournal77 (1993)
list and familiarize themselves with any new tion. Their instructorcontinued with normal
words. The interpretationtask was always ad- topicsand activitiesoutlined in the course sylla-
ministeredbefore the productiontask. bus and was unaware of the experimental in-
After the pretest was administered,classes structionoccurringwiththe other two groups.
were randomlyassigned to one of threeinstruc- For all three groups, experimentationand test-
tional treatments:traditionalinstruction,pro- ing took place in the subjects' regular class-
cessinginstruction,and no-instruction, thislast rooms during theirregular class hours.
being the control group. Subjects in the first ScoringProcedures.Both raw scores and gain
twoinstructionaltreatmentgroups received in- scores were calculated for use in two different
structionduring classtimeas part of theirnor- sets of statisticalanalyses (see below). Raw
mal routine. Both classtime and homework scores were calculated in the following way:
were matched for both traditional and pro- withrespectto the interpretationtask,each cor-
cessinginstructionin termsof timeon task.For rect response to the ten test items was given a
both groups, explicit instructionconsisted of score of one for a possible totalof ten. Correct
two consecutiveclass days with no homework. responses consisted of correctlymatchingthe
All experimentalinstructionwas performedby sentenceheard to a drawingseen. Incorrectre-
the same person, one of the researchers who sponses received a score of zero. Since the im-
was well versed in traditionalinstructionbut pact of instructionin a study of this kind is
had nopriorexperience or exposuretotheconceptof measured by an increase in knowledge gained,
processinginstruction.She was not the subjects' we decided to eliminate certain subjects from
regular classroom instructorand she did not the beginningif theydemonstrateda tendency
instructthem other than on the two days dur- to not use the word order strategyon the pre-
ing which the instructionaltreatmentwas car- test. In other words, we decided to eliminate
ried out. We should point out that at the time those subjectswho had littleto gain fromsuch
of the experimentation,thisinstructorbelieved instruction.An arbitraryscore of eight out of
thattraditionalinstructionwould resultin bet- ten on the interpretationtask was set. Anyone
ter production and that processinginstruction witheightor above on the pretesttaskforinter-
would result in better comprehension. No at- pretationwas eliminatedfromthe study.
temptwas made to swayher fromher hypothe- For the productiontask,raw scores were cal-
ses (cf.the resultspresentedlaterin thispaper). culated by counting each correct response to
At the end of the second day of instruction, the fivetestitems.Responses were givena score
subjects were given the firstpost-test.Those of two pointsif the subjectsproduced a correct
subjectswho received version A for the pretest directobject cliticpronoun formin the correct
received version B as the firstpost-test.Those position within the sentence. Responses were
who received versionB for the pretestreceived scored zero points if the subjects produced no
versionA as the firstpost-test.Versions C and cliticpronoun format all (even if the sentence
D were administeredone week afterinstruction was correctbyall otherstandards). For all those
and again one month later. As in the case of cases in between, a liberal scoring procedure
the pretestand the firstpost-test,those subjects was adopted. Responses were given a score of
who received versionC for the second post-test one point if the subjectsproduced eithera cor-
receivedversionD as the thirdpost-test.Those rectdirectobject cliticpronoun formin the in-
who received versionD forthe second post-test correct position, an incorrect clitic pronoun
received version C as the third post-test.The formin a correctposition,or an incorrectclitic
administrationof a post-testone month after pronoun form in an incorrect position. The
instructionwas to determinewhetheror not in- two, one, zero scoring procedure was done
structionhad more than an immediate impact since an either/orscoringprocedure would not
on a learner'sdeveloping system.Subjects must reveal possible intermediateeffectsof instruc-
have been presentat two of the three post-tests tion.That is, instructioncould have had an im-
(in addition to the pre-test)in order to be in- pact on the learner'sdeveloping systembut not
cluded in the study. The scores of those who necessarilyresultingin native-likeaccuracy or
had missed one post-testwere entered as miss- use. Since there were fiveitemswitha possible
ing values. score of two each, the productiontask was also
While the experimentationwas carried out worthten points. As in the interpretationtask,
in the traditionaland processing groups, the subjectswere eliminatedfromthe studyif they
no-instruction (control)group had regularclass scored eight points or above on the pretest.
hours and did not receive any special instruc- Gain scoreswere calculated bycomparingthe
Bill VanPattenand TeresaCadierno 51
raw scores on each post-testwiththe raw scores
on the pretest.Thus, therewere threedifferent FIGURE IV
Results of ANOVA with Repeated Measures Using
gain scores forthe interpretationtestand three
differentgain scores for the production test. Raw Scores on the InterpretationTest
Analysis.Raw scores were submittedto two
separate one-way analyses of variance (AN- -] Pre-test
0
OVA) with a repeated measures design. The Q Post-test I
firstANOVA was conducted on the interpreta- 9
E Post-test 2
tion data and the second on the production
Post-test 3
data. Gain scoreson the interpretationtestwere
7
submittedto three one-wayANOVAs, one for
the gains measured in each post-test.Likewise, 6
for the production task, the gain scores ob-
5
tained foreach of the threepost-testswere sub-
mittedto three one-way ANOVAs.
In all cases, the independent variable was
typeof instruction,whichconsistedof threelev-
els: traditionalinstruction,processing instruc-
tion and no instruction.
No Instruction Processing Traditional

RESULTS & DISCUSSION

An ANOVA conducted on the pretestsalone
revealed no differencesbetween the groups be-
fore instruction(p = .79 for interpretation TABLE III
task; and p = .83 for production task). We are Mean Gain Scores on InterpretationTask
thusconfidentthatany comparativeeffectsdue
No
to instructionare not related to prior knowl-
instruction Traditional Processing
edge or abilityof any one group.
The ANOVA with repeated measures con- Post-test1 1.6 .8 5.5
ducted on the raw scores of the interpretation Post-test2 2.6 .4 4.4
tasks (see Table II) revealed a significantmain Post-test3 1.4 2.2 5.2
effectfor instruction(p = .01), a significant
main effect for test (pre- vs. post-tests,with
p = .0001), and a significantinteractionbe- cant differencebetween traditionaland no in-
tweeninstructionand test(p = .0001). The re- struction(p = .96). In short, processing in-
sults of the ANOVA are displayed graphically struction was superior to the other two
in Figure IV. A post-hoc Sheff&test revealed instructionaltypes vis-a-visthe interpretation
that the effectfor instructionwas due to the test.
followingcontrasts:processing better than no The mean gain scores (interpretationtask)
instruction(p = .0178); processingbetterthan for each group at each post-testingtime are
traditionalinstruction(p = .0369); no signifi- listed in Table III. For the firstpost-test,the
ANOVA yielded a significantmain effectfor
instruction(p = .0001). A Sheff6post-hoctest
TABLE II revealed that the effectwas due to the mean
Mean RawScoresforInterpretation
Task gain scores of the processinginstructiongroup
being significantly differentfrom those of no
No instruction(p = .0001), as well as being signifi-
instruction Traditional Processing cantlydifferentfromthe gain scores of the tra-
Pretest 1.9 2.6 2.5 ditional group (p = .0001).
Post-test1 4.5 4.6 9.4 For the second post-test,the ANOVA yielded
Post-test2 5.6 3.9 7.5 anothermain effectfor instruction(again, p =
Post-test3 4.2 6.0 8.8 .0001). A Sheff6post-hocrevealed that the ef-
Note: For this and all tables, scores are rounded off fectwas due to the scores of the processingin-
to one decimal place. Slight inconsistenciesmay struction group being significantlydifferent
thus appear as tables are compared. fromthose of the traditionalinstructiongroup
52 TheModernLanguageJournal77 (1993)

(p = .0001). In addition, the scores of the no

instructiongroup were significantlydifferent FIGURE V
fromthose of the traditionalinstructiongroup Results of ANOVA withRepeated Measures Using
Raw Scores on the ProductionTest
(p = .02).
For the third post-test,the ANOVA yielded
anothermain effectforinstruction(p = .0002).
O Pre-test
The Sheff6post-hocrevealed the same source I
Post-testI
9
of the effectas in the firstpost-test:processing
Post-test 2
instructionsignificantly differentfrom no in- 8*
VA

Post-test 3
struction(p = .0004), and processing instruc-
7
tion significantlydifferent from traditional
(p = .0064). 6
The answer to our firstresearchquestion ap-
pears to be thatprocessinginstructionhas some
effecton the developing systemof language V e ~/
.-/i~
learners of Spanish as far as using the system eiv
No n r t n r e i Ta t n
to make correct form-meaning connections
during input processing. In addition, we can
partiallyanswer the third research question:
some differentialeffectexistsforprocessingin-
struction since traditional instructionappar-
ently did little to improve learners' making
form-meaningconnectionswhile processingin-
put. In order to answer our second question
and to completely answer the third, we now TABLE V
turnour attentionto the resultsof the produc- Mean Gain Scores on ProductionTask
tion task.
No
The ANOVA with repeated measures con-
instruction Traditional Processing
ducted on the raw scores of the production
tasks(see Table IV) revealed a significantmain Post-test1 2.1 5.6 6.7
effectfor instruction(p = .0119), a significant Post-test2 2.8 4.3 5.9
Post-test3 2.1 4.7 6.1
main effect for test (pre- vs. post-tests,with
p = .0001), and a significantinteractionbe-
tween instruction and test (p = .0077).
ANOVA results are displayed graphically in cessinginstructionon the productiontask,and
Figure V. A post-hoc Sheff6testrevealed that on the second and third post-tests,the raw
the effectfor instructionwas due to only one mean scores between these two groups were
contrast:traditionalbetterthan no instruction roughlythe same.
(p = .0167). No significantdifferencesbetween As in the interpretationtask,we also submit-
the processing and no instructiongroups ob- ted gain scores on the production tasks to a
tained,but more importantly, theanalysisdidnot statisticalanalysis.The mean gain scores on the
yieldanysignificantdifferencebetween in-
processing productiontask based on all three post-testare
structionand traditional In short,tra-
instruction. given in Table V. The ANOVA on the firstset
ditional instructionwas not superior to pro- of gain scores yielded a main effectforinstruc-
tion. A post-hocSheff6revealed thatthe effect
was due to the low scores of the no instruction
TABLE IV group and not to any differencesbetweentradi-
Task
Mean RawScoresforProduction tionaland processinginstruction(traditionalvs.
no instruction,p = .02; processing vs. no in-
No struction,p = .004; traditionalvs. processing,
instruction Traditional Processing
p = .74).
Pretest 2.1 3.6 1.8 For the gain scores based on the second post-
Post-test1 4.2 9.3 8.3 test,the ANOVA did not yieldany main effects
Post-test2 4.2 7.9 8.0
(p = .12). A post-hoc Sheff6revealed that all
Post-test3 4.2 8.3 8.3
comparisonswere not significant.
Bill VanPattenand TeresaCadierno 53
For the gain scores based on the third post- that learners may develop two systems-an
test,the ANOVA yielded a significantmain ef- acquired competence and a learned compe-
fectfor instruction(p = .03) and the post-hoc tence-and has claimed thattraditionalinstruc-
Sheff&revealed thatthe effectwas due entirely tion resultsin learned competence,but onlyby
to processing instruction vs. no instruction accessing comprehensible input can the ac-
(p = .04). No differenceobtained for tradi- quired systembuild up. Likewise,Schwartzhas
tional vs. no instruction(p = .17) or for tradi- suggested that the language module in the
tional vs. processing(p = .68). mind can only operate on primary linguistic
Thus the answer to the second researchques- data, and thatexplicitpracticeand negativeevi-
tion appears to be that processing instruction dence are not usable by the module. Explicit
does have an effecton production, at least in practice and negative evidence can result in
the way that we tested production. Given the what she calls LLK ("learned linguisticknowl-
answer to thisquestion, the answer to the third edge"), a systemdistinctfrom the underlying
research question is that processing and tradi- competence used by the language module.
tional instructionapparentlydo not impact on While Krashen and Schwartz may or may not
the learner in the same manner. What appears agree witha focus on input as a teachingstrat-
to have happened in this study is that pro- egy, the data in our study do support their
cessinginstructionaltered the way in whichthe claims regardingacquisition.
subjectsprocessed input, which in turn had an An anonymous reviewer of this paper has
effecton the developing systemand what the suggestedthatwe need not posit a dual knowl-
subjectscould access for production. The con- edge systemto explain the differentialeffects
verse was the not the case for the subjects in of instructionin the presentstudy,thatour re-
the traditionalgroup. That is, while traditional sults can be traced to the fact that the pro-
instructionapparently had an impact on what cessing group received "more information"
the subjectscould access for production,it had about Spanish than did the traditionalgroup.
littleimpact on how the subjects processed in- Essentially,that is, in the traditional group,
put. When compared to the no instruction OVS sentenceswere not practiced,while in the
group, the slightgains made over time by the processing group OVS sentences were part of
subjectsin the traditionalgroup cannot be at- the inputand were "explained" during the pre-
tributedto instruction. sentationon object pronouns. Since the inter-
pretationtask consisted of OVS sentences,the
DISCUSSION processinggroup was favored.While thisexpla-
nation is certainlypossible, it seems unlikely.
The resultsreported in the previous section One would need to believe thatthe production
are, to say the least, interesting.While the pro- practice of OV sentences in the traditional
cesses depicted in Figures I and III help us to group and success in producing these strings
understand why processing instructionwould on the production task did not carry over to
have an effecton production,we have no expla- interpretation of OVS sentences. In other
nation forwhytraditionalinstructionwithout- words, one would have to believe that subjects
put had littleeffecton interpretation.At this in the traditionalgroup could successfullyin-
point, we see only two possible explanations. terpretOV stringsbut not OVS strings.This
The firstis that the subjects in the production assertion is counterintuitiveand is not sup-
group learned to performthe task but did not ported by previous research in which learners
acquire any new language. However, thisexpla- of Spanish interpretedOV and OVS sentences
nation is problematic. In order to perform a in the same manner (see 38 and 22). That is,
language task, one must have some kind of regardless of whether a subject noun(phrase)
knowledge.That is, to produce a sentence,even followsa verb or not, learners of Spanish tend
in a mechanical sense, one must draw upon to process the object cliticpronouns as subject
some knowledge source to put together that pronouns.
sentence.We believe thereforethatanotherex- We should state that we are well aware of
planationis more tenablethan the first;namely, some of the methodological objections that
that traditional grammar presentation and could be raised regardingthe study.One objec-
practicedo not feed into the developing system tion is that the interpretationtask was similar
directlybut instead resultin a differentknowl- to some of the activitiesin which the learners
edge system.Krashen (20; 21) has suggested in the processing group engaged during in-
54 TheModernLanguageJournal77 (1993)
struction.It could be argued thatthe use of this duction sentencesto see if the sentencesscored
task biased the outcome toward the processing as "one" were qualitativelydifferentbetween
group on the interpretationtask. Recall, how- the two groups. We could not discern any dif-
ever, that the production task was added as a ference.
measure specificallyagainst such a bias. Since Our final comment concerns the linguistic
at no time during instructionwere subjects in item used in this study. Given that we define
the processinggroup asked to produce a single input processingas makingform-meaningcon-
sentence in which either word order was non- nections during comprehension and that the
SVO or contained a cliticobject pronoun, we resultantconnectionsare intake (i.e., are avail-
thinkthatthe overall resultsare striking.On a able foraccommodationby the developing sys-
production task not related to interpretation, tem), it can rightfullybe asked: what of those
the processing group was as competent as, if itemsthat carryno referentialmeaning? How
not betterthan, the traditionalgroup. do theyfigureinto input processing? It is im-
We now turn to a related potentialobjection portant to remember that SLA is more than
to the study. While we may have shown that just input processing. It consistsalso of those
instructionin processinginput transfersto pro- processes in Figure I that are labeled as II.
duction under controlled conditions,we have Grammars accommodate new intake and re-
not shown thatit transfersto productionunder structureas partof theirdeveloping nature.We
more spontaneous circumstances.We acknowl- suggestthatmany non-meaningbearing forms
edge thisobjection,but offerthe followingar- and structuresare firstprocessed in the input
gument. We did not set out to testwhetheror as part of somethingthat does carrymeaning.
not instructionresultsin bettercommunicative These are then storedin the developing system
performance.We set out to ascertain whether until such a time when relevantdata reveal to
or not instructioninvolvinga focus on input the learnerthatthe itemsneed to be segmented
processing resulted in similar or differentef- offor reanalyzed(29; 36). A detailed discussion
fectscompared to traditionalinstructionunder of this is not possible here, but at present we
normal classroom circumstances. We believe do not see thatnon-meaningbearingitemspose
thatwe have accomplished thiscomparison. As a problem foreitherthe resultsobtained in the
part of a series of studies on this question, we present study nor for continued research on
will examine samples of spontaneous speech the role of input processing in SLA and pro-
and writinggathered before and afterinstruc- cessinginstructionitself.
tion to see whetheror not an effectof the two
instructionaltypes on more spontaneous per-
CONCLUSION
formancecan be found.
One last possible methodologicalobjectionto In this paper we have attemptedto make a
this study (and others like it) is that its essen- connection between input processing and in-
tiallyquantitativenatureobscures possiblequal- struction.Given the role placed upon input in
itative differences on the production task. SLA, instructionas direct interventionon a
Given the scoring procedure of two points for learner's strategiesin input processing should
a well formedsentence,one point for a variety have a significanteffecton the learner's devel-
of non-native-likesentences, and zero points oping system.Resultsof the presentstudysup-
for sentences that basically reveal no learning portthisclaim. Furthermore,we have been able
at all, the question could be raised whetheror to show thatinstructionis apparentlymore ben-
not the nonsignificantdifferencebetween the eficialwhen it is directed toward how learners
traditionaland processinggroups masksdiffer- perceive and process input rather than when
ent typesof output errorsor productionstrate- instructionis focused on having learners prac-
gies. That is, one group could have a prepon- tice the language via output. Learners who re-
derance of twosand zeros whilethe othera slew ceive instructionthat attempts to alter input
of ones. In order to address this objection we processingreceive a double bonus: betterpro-
conducted a multipleANOVA using the num- cessing of input and knowledge that is appar-
ber of twos, ones and zeros received by each entlyalso available for production.The results
subject in each instructionalgroup as the de- are important,then, not only because of what
pendent measure. The resultsdid not yieldany they mightcontributeto the on-going discus-
main effect;a post-hocSheff6testrevealed no sion of the effectsof instructionbut also forthe
differencesbetween any of the possible group support that theygive to input processingas a
comparisons.In addition,we looked at the pro- criticalaspect of classroom SLA.
Bill VanPatten and Teresa Cadierno 55

sion of this paper. We would also like to thank Patsy

NOTES Lightbown,Nina Spada, and Lydia Whiteforinviting
us to present our research at the special colloquium
"The Role of Instructionin SLA" held at Concordia
1We thank Alice
Omaggio Hadley, Stephen University,Montreal, July 1991. Responsibilityfor
Krashen, Nina Garrett,and the anonymousMLJ re- the contentof the paper rests solelywithus.
viewersforcommentsand reactionsto an earlier ver-

BIBLIOGRAPHY Language Acquisition."LanguageLearningSpe-

cial Issue 4 (1976): 39-57.
15. Higgs, Theodore V. & Ray Clifford."The Push
1. Arana, Alice A. & Oswaldo Arana. Workbook toAc- Toward Communication." Curriculum,Compe-
company Puntosde Partida. New York: Random tenceand theForeignLanguageTeacher.Ed. The-
House, 1985. odore V. Higgs. Lincolnwood, IL: National
2. Bates, Elizabeth,Brian MacWhinney,CristinaCa- Textbook, 1982: 57-79.
selli,Antonella Devescovi, Francesco Natale & 16. Hilles, Sharon. "Interlanguageand the Pro-Drop
Valeria Venza. "A Cross-LinguisticStudy of Parameter."SecondLanguageResearch2 (1986):
the Development of Sentence Interpretation 33-52.
Strategies." Child Development 55 (1984): 17. Hulstijn,Jan H. "Implicitand Incidental Second
341-54. Language Learning: Experimentsin the Pro-
3. Bever, Thomas G. "The Cognitive Basis for Lin- cessingof Natural and PartlyArtificialInput."
guistic Structures."Cognitionand theDevelop- InterlingualProcessing.Ed. Hans W. Dechert
mentofLanguage. Ed. J. R. Hayes. New York: & Manfred Raupach. Ttibingen: Narr, 1989:
Wiley, 1970: 279-362. 49-73.
4. Chaudron, Craig. "A Method for Examining the 18. Klein, Wolfgang. Second Language Acquisition.
Input/IntakeDistinction."Inputin SecondLan- Cambridge: Cambridge Univ. Press, 1986.
guageAcquisition. Ed. Susan M. Gass & Carolyn 19. Knorre, Marty,Thalia Dorwick, Bill VanPatten
G. Madden. Rowley, MA: Newbury House, & Hildebrando Villarreal.Puntosde Partida.An
1985: 285-300. Invitationto Spanish. 3rd ed. New York: Ran-
5. Ellis, Rod. InstructedSecond Language Acquisition. dom House, 1989.
Cambridge, MA: Blackwell, 1990. 20. Krashen,Stephen D. Principlesand Practicein Sec-
6. . "Can Syntax be Taught? A Study of the ond Language Acquisition.Oxford: Pergamon,
Effectsof Formal Instructionon the Acquisi- 1982.
tion of WH Questions by Children." Applied 21. . "Adult Second Language Acquisitionand
Linguistics5 (1984): 138-55. Learning." SecondLanguageAcquisition and For-
7. Ervin-Tripp, Susan M. "Is Second Language eign Language Teaching.Ed. Rosario Gringas.
Learning Like the First?" TESOL Quarterly8 Washington: CAL, 1978.
(1974): 111-27. 22. Lee, James F. "Morphological Factors Influenc-
8. Faerch, Claus & Gabriele Kasper. "The Role of ing Pronominal Reference Assignment by
Comprehension in Second-Language Learn- Learners of Spanish." Language and Language
ing." AppliedLinguistics7 (1986): 257-74. Use: Studiesin Spanish.Ed. Terrell A. Morgan,
9. Felix, Sascha. "The Effectof Formal Instruction James F. Lee & Bill VanPatten. Lanham, MD:
on Second Language Acquisition." Language Univ. Press of America, 1987: 221-32.
Learning31 (1981): 87-112. 23. LoCoco, Veronica. "Learner Comprehension of
10. Fodor, Jerry A. Modularityof the Mind. Cam- Oral and WrittenSentences in German and
bridge: MIT Press, 1983. Spanish: The Importance of Word Order."
11. Garrett, Nina. "The Problem with Grammar: ForeignLanguage Learning:A ResearchPerspec-
What Kind Can the Language Learner Use?" tive.Ed. Bill VanPatten, Trisha R. Dvorak &
ModernLanguageJournal70 (1986): 133-47. James F. Lee. Rowley, MA: Newbury House,
12. Gass, Susan M. "How Do Learners Resolve Lin- 1987: 119-29.
guisticConflicts?"LinguisticPerspectiveson Sec- 24. Long, Michael H. "Does Second Language In-
ondLanguage Acquisition. Ed. Susan M. Gass & structionMake a Difference?A Review of Re-
Jacquelyn Schachter. Cambridge: Cambridge search." TESOL Quarterly17 (1983): 359-82.
Univ. Press, 1989: 183-99. 25. McLaughlin, Barry."Restructuring."AppliedLin-
13. . "IntegratingResearch Areas: A Frame- guistics11 (1990): 113-28.
work for Second Language Studies." Applied 26. Nam, Eileen. "Child and Adult PerceptualStrate-
Linguistics9 (1988): 198-217. gies in Second Language Acquisition." Paper,
14. Hatch, Evelyn & Judy Wagner-Gough. "Ex- TESOL Annual Meeting, Los Angeles, 1974.
plaining Sequence and Variation in Second 27. Newmark,Leonard. "How Not to Interferewith
56 TheModernLanguageJournal77 (1993)

Language Learning." International Journal of 35. Terrell, Tracy. "The Role of Grammar Instruc-
AmericanLinguistics 40 (1966): 77-83. tion in a CommunicativeApproach." Modern
28. Pavesi, Maria. "Markedness, Discourse Models LanguageJournal75 (1991): 52-63.
and Relative Clause Formation in a Formal 36. VanPatten, Bill. "Input Processing and Second
and Informal Context." Studiesin SecondLan- Language Acquisition:On the Relationshipbe-
guage Acquisition8 (1986): 38-55. tween Form and Meaning." Festschrift for Tracy
29. Peters, Ann. "Language Segmentation: Op- David Terrell.Ed. Peggy Heshimipour,Ricardo
eratingPrinciplesforthe Perceptionand Anal- Maldonaldo & Margaret van Naerssen. New
ysisof Language." The Cross-Linguistic Studyof York: McGraw-Hill,forthcoming.
Language Acquisition.Ed. Dan I. Slobin. Hills- 37. . "Attendingto Form and Content in the
dale, NJ: Erlbaum, 1985: 1029-67. Input." Studiesin SecondLanguageAcquisition 12
30. Pica, Teresa. "Adult Acquisitionof English as a (1990): 287-301.
Second Language under DifferentConditions 38. - . "Learners' Comprehension of CliticPro-
of Exposure." Language Learning 33 (1983): nouns: More Evidence for a Word Order
465-97. Strategy."HispanicLinguistics1 (1984): 57-67.
31. Pienemann,Manfred. "PsychologicalConstraints 39. - , James F. Lee, William R. Glass & Donna
on the Teachability of Languages." Firstand Deans Binkowski.Manual toAccompany iSabias
SecondLanguageAcquisition Processes.Ed. Carol que...? New York: McGraw-Hill,1992.
Pfaff.Cambridge,MA: NewburyHouse, 1987: 40. - ,JamesF. Lee, Terry L. Ballman, & Trisha
143-68. Dvorak. iSabias que. . .? BeginningSpanish.New
32. Schmidt,Richard. "The Role of Consciousnessin York: McGraw-Hill,1992.
Second Language Learning." AppliedLinguis- 41. White,Lydia. UniversalGrammarand SecondLan-
tics11 (1990): 129-58. guage Acquisition. Amsterdam,PA: Benjamins,
33. Schwartz,Bonnie D. "On Explicit and Negative 1989.
Evidence Affectingand EffectingCompetence 42. - . "AgainstComprehensibleInput: The In-
and Performance."Paper, Concordia Univer- put Hypothesisand the Development of Sec-
sity/McGill UniversityColloquium on the Role ond-Language Competence." AppliedLinguis-
of Instructionin Second Language Acquisi- tics8 (1987): 95-110.
tion, Montreal,July 1991. 43. Wode, Henning. "Language-Acquisitional Uni-
34. Swain, Merrill. "Communicative Competence: versals: A Unified View of Language Acquisi-
Some Roles of Comprehensible Input and tion."NativeLanguageand ForeignLanguageAc-
ComprehensibleOutput in Its Development." quisition.Ed. Harris Winitz. New York: New
Inputin SecondLanguageAcquisition. Ed. Susan York Academy of Sciences, 1981: 218-34.
M. Gass & Carolyn G. Madden. Rowley,MA:
NewburyHouse, 1985: 235-53.

APPENDIX A APPENDIX B
Sample ActivitiesUsed in Traditional Instruction Sample ActivitiesUsed in ProcessingInstruction

A. Directions: Imagine that you are in the following ActividadA. In the following,select the correctinter-
situations,performingthe indicated tasks. A friend pretationof the sentence. Keep in mind thatSpanish
asks you about particularitems.Answerlogically.Fol- has flexibleword order and does not necessarilyfol-
low the models. low subject-verb-object order like English.
1. Ud estd haciendola maletapara un viaje a Aca- 1. Me llamafrecuentemente mihermana.
pulco.(You are packing for a trip to Acapulco.) Who calls whom?
iEl trajede bafio?- iClaro que lo necesito!(The swim- a. I call my sister
ming-suit?Of course, I need it!) b. My sistercalls me
Articulos:las sandalias, las gafas de sol, los pantalones 2. Te escriben lospadres?
etc. (Articles: sandals, sunglasses,
cortos,las camisetas, Who writesto whom?
shorts,T-shirts,etc.) a. Do you writeto your parents?
b. Do your parents writeto you?
B. Directions: Rephrase sentences, changing direct 3. No nos escuchanlospadres.
object nouns to pronouns as needed. Who isn'tlisteningto whom?
1. El camarerotraelos vasosy pone los vasos en la mesa. a. Parents don't listento us.
(The waiterbringsthe glasses and puts the glasses on b. We don't listento parents.
the table.) 4. Me conocenbienmishermanos.
El camarerotraelos vasosy lospone en la mesa. Who knowswhom well?
a. My siblingsknow me.
b. I know my siblings.
Bill VanPatten and Teresa Cadierno 57

Actividad B. Listento thespeakeron thetape.Match ones applyto you. Compareyourresponseswitha

each sentenceyou hear withone of the statements classmate.
below. 1. Losllamoconfrecuenciaportelgfono.
1. A manis callingme. 2. Losvisitolosfinesdesemana.
I am callinga man. 3. Los visitopor los menosuna vez al mes.
2. Myparentsvisitme. 4. Losabrazocuandolosveo(abrazar= to
I visitmyparents. hug).
3. I am pleasingto myfamily. muybien.
5. Los comprendo
Myfamily is pleasingto me. 6. Losignoro
completamente.
4. We are greeting a friend. 7. Los
A friendgreetsus.
5. Our relatives don'tunderstand us. Did younoticethatthereareno explicit
subjectnouns
We don'tunderstand our relatives. or subjectpronounsin each sentence?Since theyo
formoftheverbcan onlyrefertoyo,no subjectpro-
E. Each sentencecorresponds
Actividad to something nounis needed.Alloftheabovesentences are ofthe
thatyou mightdo to your parents.Check which simplewordorderobjectpronoun+ verb.

The NEH Teacher-Scholar Program

THE NATIONAL ENDOWMENT FOR THE her own setting.In any case, the focus of the
Humanities offers elementary and secondary year's activityshould be intensive study on a
school teachers an opportunityto receive sup- well-definedtopic in the humanities.
port foran academic year of full-timeindepen- Teachers may apply for a teacher-scholar
dent study in history,literature,foreign lan- award if they: 1) are employed full-timein ele-
guages, and otherdisciplinesof the humanities. mentary, middle, or secondary schools and
The program is designed to give recipientsan have teaching responsibilitiesprimarilyin one
opportunityto spend a year increasing their or more of the humanitiesdisciplines-librar-
understandingof texts,topics,and issues bear- ians may apply if they spend more than fifty
ing on the subjects theyteach. percentof theirtime directlyteaching human-
Recipientsare selected in a single,nationwide ities courses; 2) have completed at least three
competition.An award is intended to replace years of full-timeteaching at the time of the
fullythe recipient'sacademic-yearsalary or to application and intend to return to teaching
supplement other grants and sabbatical pay forat least two yearsaftercompletingthe proj-
up to the amount of the academic-yearsalary. ect; 3) are US citizensor foreignnationalswho
The stipend may not exceed $30,000. Each have been residentsof the United States or its
Teacher-Scholar award offersa $500 honorar- territoriesfor at least three years immediately
ium for a mentor should the recipientelect to preceding the time of application. American
work with an expert in the field of his or her citizenswho are full-timeteachersof American
study. studentsabroad may also apply; 4) are not past
The Teacher-Scholar Program focuses on recipientsof the NEH Teacher-Scholar award.
rigorousacademic studyemphasizingexamina- Guidelines and application forms for the
tionof significantissues and importantprimary NEH Teacher-Scholar Program are available
and secondary works central to one or more fromthe Endowment.Applicationsmustbe re-
of the humanitiesdisciplines. A recipientmay ceived no later than May 1 for grant periods
establish a working relationship with nearby beginningin September of the followingyear.
scholars,academic libraries,archives,or muse- To request these guidelines and forms,please
ums. A recipientmayenroll in up to twocollege write or call: Teacher-Scholar Program/Divi-
courses or other educational programs during sion of Educational Programs Room 302/Na-
the tenure of the award, provided such course tional Endowment for the Humanities/1100
work is directlyrelevant to the proposed proj- Pennsylvania Avenue NW/Washington, DC
ect. Or, a recipient may work alone in his or 20506; phone (202) 786-0377.