Changing Theories of Language teaching

Putting Error
Correction
into
Proper Perspective
Georgette Ioup
Department of English
University of New Orleans
Changing Theories of Language teaching
Stage 1: Grammar Translation
 Stage 2: Audio Lingual Method
(ALM)
 Stage 3: a.Communicative Competence

(W. Rivers, 1968)
b. L1 Acq. = L2 Acq.
(Dulay & Burt, 1974)
Krashen’s Monitor Model
Five Hypotheses
Acquisition / Learning Hypothesis
 Natural Order Hypothesis
 Monitor Hypothesis
 Input Hypothesis
 Affective Filter Hypothesis

Acquisition / Learning


Acquisition: Subconscious processes
Learning: Conscious formal knowledge
Formal
Learning
Natural
Input
Learned
Grammar
Acquired
Grammar
Monitor
Language
Output
Input Hypothesis
Relates to Acquisition only.
 We acquire by receiving
Comprehensible input – input that is
a little beyond our current level.

Current Level – i
 Comprehensible Input - i + 1

Critical Responses to Krashen’s Theory

Higgs & Clifford, 1982
A. Terminal 2s:
Street learners with
uncorrected errors
B. Advocates “accuracy first’
(See also, Valette, 1991)
Critical Responses to Krashen’s Theory

White, 1987
Certain types of errors need rule instruction.
L1 French – L2 English
1. Mary studied the lesson carefully
2. Carefully Mary studied the lesson
3. *Mary studied carefully the lesson
Critical Responses to Krashen’s Theory
 Long,
1991
Focus on Form, not Focus on
Forms
Focus on Forms: Teaching isolated rules
Focus on Form: teaching rules in context
Critical Responses to Krashen’s Theory
 Ellis,
1990
Forms that contribute little to
communication tend to be ignored.
They need attention.
Critical Responses to Krashen’s Theory
 Schmidt,
1994
Must notice the gap.
When there is a discrepancy between
a TL form and an IL form, adults
won’t notice it unless attention is
drawn to it.
In Support of Krashen’s Theory

VanPatten, 1988
A. No evidence that “terminal learners
are terminal.”
B. Early focus on grammar will inhibit
developing fluency.
In Support of Krashen’s Theory
 Schwartz,
1993
LAD can only accept natural input
• Includes syntactic rules of Universal
Grammar
• Excludes morphology and lexicon
These two can be learned formally.
In Support of Krashen’s Theory
Paradis, 1994
Explicit Knowledge -> Declarative Memory
– Consciously accessible knowledge
– Facts and Concepts
Ex: Formal knowledge of grammar rules
In Support of Krashen’s Theory
Paradis, 1994
Implicit Knowledge -> Procedural Memory
– Information derived unconsciously
– Skills and Processes
Ex: Natural language acquisition
In Support of Krashen’s Theory
Paradis, 1994


Declarative memory cannot translate
into procedural memory.
Each uses a different part of the brain.
Sounding Alarms
The American Scene
Valette, 1991, Modern Language Journal
The level of foreign language proficiency has
deteriorated in the last 25 years. The
median proficiency score for foreign
language majors is now probably no higher
than 1+.
Two causes: “contact with inaccurate models
and the acceptance (or non-correction) of
inaccurate speech production.”
Sounding Alarms
The British Scene
Mitchell, 2000, Applied Linguistics
“Grenfell (1999) has documented current
disillusionment and uncertainty among the
MFL professional community in England
with the present National
Curriculum/GCSE framework.”
“…despite the focus on communication, a
disappointing proportion of pupils are
making the transition to creative control
of the TL system.”
Research Comparing Implicit and Explicit
Language Teaching
Scott, 1989
Explicit condition performed better
Alenan, 1995
The groups who received rules performed better
Robinson, 1995
Instructed groups better on easy rules
VanPatten & Oikkenon, 1996
Explicit rule group did worse
Conflicting results could be due to monitoring
Research on Error Correction in Writing
It is effective – improves writing
proficiency
– Cardell & Carno (1982)
It gives no advantage over no correction
– Fathman & Walley (1990)
It is ineffective – doesn’t improve writing
proficiency
– Kepner (1991)
– Semke (1984)
It impedes proficiency development
– Sheppard (1992)
Research on Error Correction in Writing

Truscott, 1996
A. Agrees with Krashen and
Schwartz
According to learning theory, structures
learned through error correction can’t
become part of the internal grammar.
B. Advocates an end to grammar
correction in writing
It is ineffective and harmful.
Research on Error Correction in Writing
Ferris (1999)
*Argues for a role for grammar correction.
*Claims studies are inclusive.
*Truscott’s radical stance is premature.
Weakness in Focus-on-Form Research
My problem with focus-on-form studies:
They cover a short
Similar view expressed in
DeKeyser, 1977
period of time.
Two Questions
1. Can formal instruction help, if
formal learning cannot become
part of the acquired grammar?
2. Why might it take a long time for
the formal learning to become
effective?
Answer to Question 1
Can instruction help?
Paradis (1994), DeKeyser (2002), Hulstijn (2002)
Explicit Knowledge -> Declarative Memory
– Consciously accessible knowledge
– Facts and Concepts
Ex: Formal knowledge of grammar rules
Answer to Question 1
Can instruction help?
Paradis (1994), DeKeyser (2002), Hulstijn (2002)
Implicit Knowledge -> Procedural Memory
– Information derived unconsciously
– Skills and Processes
Ex: Natural language acquisition
Answer to Question 1
Can instruction help?
Paradis (1994), DeKeyser (2002), Hulstijn (2002)


Declarative memory cannot translate
into procedural memory.
Each uses a different part of the brain.
Answer to Question 1
Can instruction help?
Anderson (1995), DeKeyser (2002), Hulstijn (2002),
Paradis (1994)


Each time one uses a metalinguistic
declarative rule,
One simultaneously builds
unconscious, implicit procedural
knowledge.
Answer to Question 1
Can instruction help?
Hulstijn, 2002


Metalinguistic knowledge doesn’t
actually transform into implicit
knowledge.
An implicit network that is separate
from and in addition to the explicit
system is independently constructed.
Answer to Question 2
Why might it take a long time?
McLaughlin (1990), DeKeyser (2001)
Complex skills start as controlled processes.
Given:
* Time
* Attention
They become automatic routines.
It is practice which accomplishes this.
Answer to Question 2
Why might it take a long time?
Shriffrin & Schneider, 1986
The development of automaticity for
generalized skills depends on high levels
of practice.
McLaughlin, 1990
“A skill must be practiced again and again
and again, until no attention is required
for performance.”
Answer to Question 2
Why might it take a long time?
Anderson, 1995
A second factor is important for
automaticity – feedback.
Jenny
My Study
Immigrated to the U.S. at age 10
Mainstreamed into regular classes, no ESL
Entered university at age 19
Many fossilized grammar errors
Required to take ESL classes
Needed 7 semesters plus one summer
tutoring to destabilize her errors
 I argue that rule isolation and error
correction destabilized her fossilized
structures






Type of Instruction Given



Some formal rule presentation
Consistent feedback on errors in written work
Revision with error correction required
Two Contributions that Instruction Provided
1. Made her aware of which forms
were incorrect
2. Gave her the necessary practice to
change her grammar
Table 1
Proficiency Scores on Initial Placement and End of Semester
Exams
MTAC
MTELP
COMP
Initial Placement
88
85
58
Intensive
90
86
78
Time 1
87
88
81
Time 2
92
87
81
Time 3
90
81
85
Time 1
87
89
89
Time 2
93
89
87
Time 3
90
90
88
Nonintensive I
Nonintensive 2
Figure 1. T-unit analyses of 200-word samples at 3 different levels.
Figure 2. Mean percent of error-free T-units per sample at 3 different levels.
Table 2
Comparison of Selected Errors by Level
Prep
Tense
Article
Verb Form
Plural-s
4
18
19
15
9
Midway 1
4
6
4
6
Exit
2
1
6
4
Entry
4
Persistent Errors
1. Because of + S
“Because of my father’s restaurant short
of hand.”
2. Make + Infinitive
“made the car to go in reverse”
Finally eradicated at exit level
Conclusion
Focus-on-form combined with
abundant contextualized
practice and feedback can aid
a communicative learning
approach.