Processing Instruction and Structured Input

1
Chapter 7

• Processing Instruction and Structured Input

2
In this chapter we explore

• The nature of input processing
• Processing instruction grammar instruction that
  has structured input at its core
• Research on processing instruction that
  demonstrates its effectiveness
• A set of guidelines for developing structured
  input activities

3
Input processing

• Traditional instruction consisting of drills in
  which learner output is manipulated and the
  instruction is divorced from meaning or
  communication is not an effective method for
  enhancing language acquisition.
• What is needed is a new pedagogy of grammar
  instruction that takes as its point of departure
  what we know about how grammatical forms and
  structures are acquired.
• This pedagogy needs to work with input and with
  the processes learners use to get data from that
  input.

4
Input processing

• Of concern is input processing, how learners
  initially perceive and process linguistic data in
  the language they hear.
• In input processing, learners might encounter
  their first problems in dealing with the
  properties of the new language.
• We must come to some understanding of what input
  processing looks like.

5
Intake from input

• Input processing is concerned with those
  psycholinguistic strategies and mechanisms by
  which learners derive intake from input.
• Intake refers to the linguistic data in the input
  that learners attend to and hold in working
  memory during online comprehension.

6
Form

• Research on input processing attempts to explain
  how learners get form from input while their
  primary attention is on meaning.
• Form here is defined as surface features of
  language, although input processing is also
  relevant to syntax.

7
The most complete model

• VanPatten (1996,2003b) presents the most complete
  model of input processing in SLA.
• The role of working memory is important in this
  model since some of the principles are predicated
  on a limited capacity for processing.
• Humans develop mechanisms that allow them to
  selectively attend to incoming stimuli. Without
  such mechanisms, there would be informational
  overload.

8
VanPattens Principles

• Principle 1(P1). The Primacy of Meaning
  Principle. Learners process input for meaning
  before they process it for form.
• P1a. The Primacy of Content Words Principle.
  Learners process content words in the input
  before anything else.
• P1b. The Lexical Preference Principle. Learners
  will tend to rely on lexical items as opposed to
  grammatical form to get meaning when both encode
  the same semantic information.
• Example I went to the store yesterday.

9
VanPattens Principles continued

• ? P1c. The Preference for Nonredundancy
  Principle. Learners are more likely to process
  nonredundant meaningful grammatical form before
  they process redundant meaningful forms.
• Examples
• I went to the store yesterday. (redundant past)
• I went to the store. (non-redundant past)

10
Principle 1c

• In this principle, more, less, and
  nonmeaningful refer to the communicative value
  that a grammatical feature contributes to overall
  sentence meaning.
• Communicative value refers to the meaning that a
  form contributes to overall sentence meaning and
  is based on two features.
• /- inherent semantic value
• /- redundancy

11
Communicative value

• A form that can be classified as having inherent
  semantic value and is not a redundant feature of
  language, will tend to have high communicative
  value.
• For example
• In English verbal morphology, -ing tends to have
  high communicative value
• 1) It has inherent semantic value because it
  encodes progressive aspect (i.e., -ing in
  progress)
• 2) -ing tends to be high in communicative
  value because it is seldom redundant in naturally
  occurring discourse since no lexical information
  in the utterance co-occurs to provide cues to
  aspect.

12
Nature of communicative value

• In order to grasp the semantic notion of in
  progress the L2 learner of English must process
  the verbal inflection ing in the input.
• The nature of communicative value is important
  for input processing.
• The more communicative value a form has, the more
  likely it is to get processed and made available
  in the intake data for acquisition (P1d).

13
Conversely

• The less communicative value a form has, the more
  likely learners are to skip it in the input.
• For learners to process forms of little or no
  communicative value in the input, they must be
  able to comprehend an utterance such that the act
  of comprehension does not tie up all their
  attentional resources.

14
VanPattens Principles continued

• ?P1d. The Meaning-before-Nonmeaning Principle.
  Learners are more likely to process meaningful
  grammatical forms before nonmeaningful forms
  irrespective of redundancy.
• Example
• Los gatos negros.
• (masculine non-meaningful)
• (plural s meaningful)

15
VanPattens Principles continued

• ? P1e. The Availability of Resources Principle.
  For learners to process either redundant
  meaningful grammatical forms or non-meaningful
  forms, the processing of overall sentential
  meaning must not drain available processing
  resources (i.e. no overload of information) .

16
VanPattens Principles continued

• ? P1f. The Sentence Location Principle. Learners
  tend to process items in sentence initial
  position before those in final position and those
  in medial position.
• ? For example, learners are much more likely
  to pick up question words and their syntax than
  object pronouns or the subjunctive, which tends
  to occur inside the sentence.
• Example Hablo español.
• Yo quiero que tú hables español.

17
An example in French

• Example Jean fait promener le chien à Marie.
  (French causative)
• Students John walks the dog for Mary.
• Correct John makes Mary walk the dog.
• (Students will incorrectly encode Jean as the
  subject of the second verb and thus delivering
  erroneous intake to their developing linguistic
  systems.)

18
VanPattens Principles continued

• Principle 2. The First Noun Principle. Learners
  tend to process the first noun or pronoun they
  encounter in a sentence as the subject or agent.
• Example A Juan no le gusta helado.
• (Students will incorrectly encode Juan as the
  subject of the sentence and thus delivering
  erroneous intake to their developing linguistic
  systems.)

19
VanPattens Principles continued

• ?P2a. The Lexical Semantics Principle. Learners
  may rely on lexical semantics, where possible,
  instead of word order to interpret sentences.
• (see previous examples)

20
Word order

• Word order is important in Input processing.
• P2, the first noun principle, may have important
  effects on the acquisition of a language that
  does not follow strict subject-verb-object (SVO)
  word order.

21
Erroneous input

• Research has shown that learners do indeed encode
  such pronouns and noun phrases as subjects, thus
  delivering erroneous input to their developing
  linguistic systems.
• They think that Juan is the subject.
• It is not that meaning is gotten elsewhere it is
  that meaning is not gotten at all or is gotten
  wrong.
• The form-meaning connections are not only
  filtered, they are altered.

22
VanPattens Principles continued

• ? P2b. The Event Probabilities Principle.
  Learners may rely on event probabilities (i.e.
  whats more likely to happen), where possible,
  instead of word order to interpret sentences.
• ? For example, fluent English speakers would
  assign the semantic role of agent to the hunger
  and the role of patient to the lion The lion
  was killed by the hunter.
• ? Research has shown that learners of English as
  an L2 incorrectly interpret it as The lion
  killed the hunter.

23
VanPattens Principles continued

• P2c. The Contextual Constraint Principle.
  Learners may rely less on the First Noun
  Principle if preceding context constrains the
  possible interpretation of a clause or sentence.
• (see previous example)

24
Summary

• Research on input processing attempts to
  describe
• What linguistic data learners attend to during
  comprehension
• Which ones they do not attend to
• What grammatical roles learners assign to nouns
• How position in an utterance influences what gets
  processed.
• Intake is grammatical information as it relates
  to meaning that learners have comprehended (or
  think they have comprehended.)

25
A reminder

• As a reminder, input processing is but one set of
  processes related to acquisition.
• That learners derive some kind of intake from the
  input does not mean that the data contained in
  the intake automatically make their way into the
  developing mental representation of the L2 in the
  learners head (i.e., intake does not equal
  acquisition).

26
Rethinking grammar instruction Structured input

• We now have some idea of what learners are doing
  with input when they are asked to comprehend it.
• We can begin to develop a new kind of grammar
  instruction-one that will guide and focus
  learners attention when they process input.

27
Processing instruction

• Processing instruction consists of three basic
  components
• Learners are given information about a linguistic
  structure or form.
• Learners are informed about a particular
  processing strategy that may negatively affect
  their picking up of the form or structure during
  comprehension.
• Learners are pushed to process the form or
  structure during activities with structured
  input- input that is manipulated in particular
  ways to push learners to become dependent on form
  and structured to get meaning.

28
Processing-oriented grammar instruction

• Input?Intake?Developing System?Output
• Processing Mechanisms
• Focused Practice

29
An example of relating processing strategies to
instruction Verb morphology

• We turn to activities that focus learners
  attention on verb endings the goal is for
  learners to use these morphological endings to
  comprehend tense rather than solely rely on
  lexical items.
• After learners receive a brief explanation of how
  past-tense endings work, they might first
  practice attaching the concept of past time to
  verb forms in an activity such as the following.

30
Listening for time reference

• Listen to each sentence. Indicate whether the
  action occurred last week or is part of a set of
  actions oriented toward the present.
• John talked on the phone.
• Mary helped her mother.
• Robert studies for two hours.
• Sam watched TV.
• Lori visits her parents.

31
Structuring the input

• Note that only the very ending encodes tense in
  the input sentence.
• Lexical terms and discourse that would indicate a
  time frame are not present, thereby encouraging
  learners to attend to the grammatical markers for
  tense.
• The input has been structured.

32
An example of relating processing strategies to
instruction Adjective agreement

• This time we focus on the following strategy
  P1d. Learners are more likely to process
  meaningful grammatical forms before nonmeaningful
  forms, irrespective of redundancy.
• Some features of language do not have inherent
  semantic or communicative value.

33
For example

• In the Romance languages, adjectives must agree
  in number and gender with the nouns they modify,
  but this feature of grammar contributes little or
  nothing to the meaning of the utterance in most
  cases.
• In the following Spanish-language activity,
  learners attention is directed toward proper
  adjective form by a task in which the adjective
  endings must be attended to.

34
Who is it?

• Listen to each sentence in which a person is
  described. Determine which person is being
  described and then indicate whether you agree or
  disagree.
• David Letterman Madonna
• Es dinámica. (Shes dynamic.)
• Es comprensivo. (Hes understanding.)
• Es reservada. (Shes reserved.)

35
An example of relating processing strategies to
instruction The French causative

• Remember that learners apply a first-noun
  strategy to determine subjects and objects of
  sentences (who did what to whom)
• With the French causative, this leads to
  misinterpretation and nonacquisition.
• In this activity, learners are pushed to process
  correctly to be sure this happens, sentences
  with the noncausative faire (faire du ski, to
  ski) that involve two people are also included.

36
Who is performing?

• Listen to each sentence. Then answer the
  question.
• Who cleans the room?
• Who packs the bags?
• Teachers script Read each sentence ONCE.
  After each sentence, ask for an answer.
• Claude fait nettoyer la chambre à Richard.
  (Structured)
• Marc fait les valises pour Jean.

37
Research on processing instruction

• There has been ongoing research regarding the
  effectiveness of processing instruction.
• An important part of this research has examined
  the relative effects of processing instruction
  versus those of traditional instruction.
• The study that launched this research agenda is
  VanPatten and Cadierno (1993).
• It is the most frequently cited study and has
  been the impetus for a number of replication
  studies.

38
Research questions

• VanPatten and Cadierno sought to answer the
  following research questions
• Does altering the way in which learners process
  input have an effect on their developing systems?
• If there is an effect, is it limited solely to
  processing more input or does instruction in
  input also have an effect on output?
• If there is an effect, is it the same effect that
  traditional instruction has (assuming an effect
  for the latter)?

39
Focus of the research

• VanPatten and Cadierno compared three groups of
  learners
• A processing instruction group (number27)
• A traditional instruction group (number26)
• A control group (number27)
• The processing group received instruction based
  along the lines presented earlier.
• The focus was word order and object pronouns in
  Spanish.

40
Who did what

• In the processing treatment, learners first
  received activities with right or wrong answers
  (Select the picture that best goes with what you
  hear) followed by activities in which they
  offered opinions.
• In the traditional group, learners received
  involving a typical explanation of object
  pronouns and the complete paradigms of the forms.
• The control group did not receive instruction on
  the target structure and instead read an essay
  and discussed it in class.

41
Assessment

• Assessment consisted of two tests a
  sentence-level interpretation test and a
  sentence-level production test.
• These were administered as a pretest, an
  immediate posttest, a two-week delayed posttest,
  and a four-week delayed posttest.

42
Assessment continued

• The interpretation test consisted of ten target
  items and ten distractors.
• The production test consisted of five items with
  five distractors.
• The interpretation group was based on an activity
  performed by the processing group (Select the
  picture that best goes with what you hear.)
• The production test was based on an activity the
  traditional group performed (Complete the
  sentence based on the pictures you see.)

43
Results!

• The pretests yielded no differences among the
  groups on the two tests prior to treatment.
• In the posttesting phase, the processing group
  made significant gains on the interpretation test
  whereas the traditional and control groups did
  not.
• On the production test, both the traditional and
  processing groups made significant gains but were
  not significantly different from each other.
• The control group did not make significant gains.

44
Conclusions

• Altering the way learners process input could
  alter their developing systems.
• The effects of processing instruction are not
  limited to processing but also show up on
  production measures.
• The effects of processing instruction are
  different from those of traditional instruction.

45
Two for one

• By being pushed to process form and meaning
  simultaneously, learners with processing
  instruction not only could process better but
  also could access their newfound knowledge to
  produce a structure that never produced during
  the treatment stage.
• Members of the traditional group learner to do a
  task, while the members of the processing group
  actually experienced a change in their underlying
  knowledge that allowed them to perform on
  different kinds of tasks.

46
Areas for future research

• Are the effects of processing instruction (PI)
  generalizable to other structures?
• Are the effects of PI due to different explicit
  information?
• Are the effects of PI observable with different
  assessment tasks?
• Are the effects of PI different from the effects
  of other types of instruction?
• Do the effects of PI hold over time?

47
Are the effects of P1 generalizable to other
structures?

• Cadierno (1995) replicated the VanPatten and
  Cadierno study using the Spanish preterite (past)
  tense as the target structure.
• Again contrasting a control group, a traditional
  instruction (TI) group, and a processing
  instruction (PI) group, Cadierno measured the
  effects of treatment via two measures
• An interpretation test (Is the sentence youre
  hearing present, past, or future?)
• A production test (writing sentences in the past)

48
Results

• Cadiernos results matched those of VanPatten and
  Cadierno exactly
• On the interpretation test, the PI group improved
  significantly, but the other two groups did not.
• On the production test, both the PI and TI groups
  improved significantly but were not different
  from each other.

49
Chengs study

• In her dissertation, Cheng (1995) conducted a
  study with ser and estar, the two major copular
  verbs in Spanish.
• She compared a control, a processing, and a
  traditional group in the use of copular verbs
  with adjectives as the target.
• Her results mirrored those of the original
  VanPatten and Cadierno study.

50
Farleys study

• In another study, Farley (2001a) demonstrated the
  effects of PI on the Spanish subjunctive with
  noun clauses.
• In his study he showed that participants who
  received PI made significant gains in both
  interpretation and production abilities with the
  subjunctive both in form and use.

51
Bucks dissertation

• Buck (2000) investigated the relative effects of
  PI and TI in the acquisition of the present
  continuous (versus the present progressive) in
  English by native speakers of Spanish.
• Bill is smoking a pipe versus Bill smokes a
  pipe.
• He results indicated greater gains for the
  processing group that were maintained over time
  on the interpretation test.

52
VanPatten and Wong

• In one other study, VanPatten and Wong (2003)
  demonstrated that PI was superior to TI with the
  French causative.
• They compared a control, a processing, and a
  traditional group and measured outcomes with an
  interpretation and a production test.
• Their results were the same as the results of the
  original study.

53
Acquisition of verbal morphology

• In another study involving the acquisition of
  verbal morphology, Benati (2001) compared PI, TI,
  and a control group using the Italian future
  tense as the target structure.
• His results were similar to but not the same as
  those of the original study.
• On the interpretation task, the PI group improved
  significantly, the TI group did as well, and the
  control group did not.
• However, the gains made by the PI group were
  significantly greater than those of the TI group.
  PIgtTIgtC

54
Are the effects of PI due to different explicit
information?

• In VanPatten and Oikennon (1996), the researchers
  compared three groups.
• One that received PI exactly as in the original
  VanPatten and Cadierno study.
• Another that received the structured input
  activities only, with no prior explicit
  information and no explanation during the
  activities
• Another that received explicit information only,
  with no structured input activities.
• The researchers used the same assessment tests as
  in the original study.

55
Results

• Both the regular processing group and the
  structured input-only group improved
  significantly but were not different from each
  other.
• The effects of PI are due not to the explicit
  information provided to learners but to the
  particular nature of the structured input
  activities.

56
Computer-assisted language learning (CALL)

• Sanz and Morgan-Short (2003) set out to test
  whether explicit feedback is necessary or helpful
  to learners.
• They tested four groups using the variables /-
  explanation and /- explicit feedback.
• All groups, regardless of the combinations of
  these variables, received the same structured
  input as practice.

57
The groups

• The first group was explanation (explicit
  information about the language and how to process
  it in the input) and explicit feedback
  (telling learners not only whether an answer is
  correct or not but what the problem is if the
  answer is not correct0
• The second group was - explanation and -
  explicit feedback (learners received structured
  input only, with indications only of whether
  their answers were right or wrong.

58
The groups continued

• The third group was explanation but -
  explicit feedback.
• The fourth group was - explanation but
  explicit feedback.

59
Results

• The results showed that all groups improved
  significantly on the three assessment tasks
  (interpretation and two production tasks- a
  sentence-completion task and a video-retelling
  task) from pre- to posttests.
• The researchers found that no group was better
  than any other on any task.
• Neither explicit information nor explicit
  feedback seemed to be crucial for a change in
  performance.
• Practice in decoding structured input alone (as
  in the second group) seemed to be sufficient.

60
Benatis study

• In one other recent study, Benati (2003) reported
  similar findings with the Italian future tense.
• He compared a regular PI groups with a structured
  input-only group and an explicit information-only
  group.

61
Benatis results

• The explicit-only information group improved
  slightly from pre- to posttest measures, but that
  both PI and structured input-only groups improved
  much more, and they improvement was not
  significantly different from each other.
• Both treatments were significantly better than
  the explicit information-only group.
• These results suggest a major if not causative
  role for the structured input activities of PI
  and only a minor role, if any, for explicit
  information.

62
Conclusions

• Structured input appears to be the causative
  variable in processing instruction.
• This means that explicit information is not
  important if the types of activities learners are
  engaged in actually push them to alter their
  processing strategies and make more or better
  form-meaning connections.

63
Two additional studies

• Farley (2003)- In Farleys study, the target item
  was the subjunctive in noun clauses in Spanish.
• Wongs study (2003) focused on negation of
  indefinite articles and partitives with avoir in
  French.
• In both studies, the learners who received
  structured input only, without any prior
  explanation of the rule, made significant gains.

64
Are the effects of PI observable with different
assessment tasks?

• In VanPatten and Sanz (1995) the researchers
  investigated the effects of PI as measured by
  three kinds of output tests.
• They compared a PI group to a control group,
  using the same materials as in VanPatten and
  Cadierno.

65
Their output tests

• Their output tests included not only a
  sentence-level test but also a question-answer
  test (based on pictures) and a video-narration
  test.
• They administered the output tests in two modes
  written and oral.
• In the video narration, participants must provide
  all vocabulary, all syntax, and all grammatical
  features on their own, without any prompts.

66
Their results

• VanPatten and Sanz found that the control group
  did not improve on any tests.
• The PI group improved significantly on the
  interpretation test and on the sentence-level
  test in both modes.
• Their gains were significant in the written mode
  but just missed significance in the oral mode.
• In all tests, the PI participants performed
  better on the written tests than the oral.
• It appears that the effects of PI are observable
  in a variety of output tests and are not limited
  to sentence-level tests.

67
Are the effects of PI different from the effects
of other types of instruction?

• Farley (2001a) compared the relative effects of
  PI with the effects of meaning-based output
  instruction (MOI).
• Farley based the PI materials on P1b of
  VanPattens model.
• Farleys initial activities pushed learners to
  attend to subordinate clauses without main
  clauses in Spanish and had them indicate what
  the possible main clause could have been (or vice
  versa).
• Farleys activities had learners combine main and
  subordinate clauses to express doubt and belief
  about various people, places, and events.

68
MOI (meaning-based output)

• Unlike TI (Traditional Instruction), MOI
  (meaning-based output) contains no mechanical
  drills and is based on the tenets of structured
  output activities that were first mentioned in
  VanPatten and Cadierno (1993).
• Participants might have heard (translated from
  Spanish) I dont think that dogs and on a
  sheet of paper would see (to be) intelligent.
• They would then have to indicate what the person
  might be saying by using the correct verb form.

69
Farleys procedure

• Both the PI (Processing Instruction) and the MOI
  (Meaning Based Output Instruction) groups had two
  days of instruction on the Spanish subjunctive.
• Farley assessed outcomes using a pretest and
  posttest design, with one posttest administered
  one month after treatment.
• The tests consisted of an interpretation test
  based on the PI materials and a production test
  based on the MOI materials.

70
Farleys results

• Farleys results differed from the results of the
  previous studies comparing PI with TI.
• His results showed that the PI and MOI groups
  improved significantly on both the interpretation
  and the production tests, with no difference
  between them.
• Thus, PI was not superior to MOI, and neither was
  MOI superior to PI.

71
Another Farley study

• In Farley (2001b), he used the same design,
  procedure, and target structure as in his 2001a
  study.
• His results Although both groups improved on the
  interpretation task, only the PI group maintained
  its performance on a delayed task.
• The MOI group declined in performance.
• Thus, PI did prove to be superior to MOI in the
  long run.

72
Do the effects of PI hold over time?

• In the studies reported so far, the longest time
  delay in administering a posttest in any one
  study was four weeks.
• As we saw with Lightbowns 1983 study, after one
  year the effects of instruction wore off and
  learners were back where they had been at the
  beginning.

73
Long-term effects of PI

• In VanPatten and Fernández (2003), the long-term
  effects of PI were studied.
• An immediate posttest was given after
  instruction, and another one was given eight
  months later to the students who had continued on
  to the next semester and who had completed all
  phases of study.

74
Results

• When VanPatten and Fernández compared the
  posttest results to the pretest results, they
  found that, as in all other studies, the students
  improved significantly on both measures.
• Even though the scores dropped somewhat on the
  eight-month delayed test, the students were still
  significantly better at performing the tests than
  they were prior to treatment.
• At least in this one study, the observed effects
  of PI seem to be durable.

75
Guidelines for developing structured input
activities

• Present one thing at a time.
• Keep meaning in focus.
• Move from sentences to connected discourse.
• Use both oral and written input.
• Have the learner do something with the input.
• Keep the learners processing strategies in mind.

76
Present one thing at a time

• Structured input must be delivered to the
  learners developing system in an efficient way.
• Maximum efficiency is achieved when one function
  and one form are the focus at any given time.
• Breaking up paradigms and lists of rules if
  useful for two reasons.
• It allows the explicit presentation and
  explanation of the grammatical structure to be
  kept to a minimum.
• Breaking up a paradigm is more likely to result
  in attention directed toward the targeted item.

77
Keep meaning in focus

• Learners should not engage in the mechanical
  input activities of traditional grammar
  instruction.
• Remember that input should be attended to for its
  message so that learners can see how grammar
  assists in the delivery of that message.

78
ActivityLooking for verb endings

• Check off the statements you think are true
  based on what you know about your instructor.