Skill Acquisition Theory

Skill Acquisition Theory (SAT) is a cognitive theory of learning, applied to SLA primarily by Robert DeKeyser. It draws on Anderson's (1982, 1993, 2007) ACT (Adaptive Control of Thought) framework and proposes that language learning follows the same path as learning any complex cognitive skill — from driving a car to writing computer programs.

The Three Stages

1. Declarative stage — learner acquires explicit knowledge of rules ("knowing that")
   • Knowledge ABOUT a skill, acquired through observation, verbal transmission, or rule presentation
   • e.g., "In English, the past tense of regular verbs adds -ed"
2. Procedural stage — through practice, declarative knowledge becomes procedural ("knowing how")
   • Proceduralisation can be complete after relatively few trials (DeKeyser, 1997: after just 16 practice items)
   • Procedural knowledge is a "ready-made chunk" — no longer requires assembling declarative pieces
3. Automatic stage — further practice leads to automatisation
   • Automaticity is not all-or-nothing — even highly automatised behaviours can break down under pressure
   • Follows the power law of learning: reaction time and error rate decrease following a power function
   • Requires massive amounts of practice ("overlearning") beyond initial mastery

Key Properties from DeKeyser & Suzuki (2025)

Skill specificity

Procedural knowledge does not transfer well — even to seemingly similar tasks. Comprehension training doesn't significantly reduce production errors; production training doesn't hinder comprehension. This means practice must be task-specific, not generic (De Jong, 2005; Li & DeKeyser, 2017).

Implication for PPP: A "Produce" phase that asks learners to use the target form in a different mode (e.g., writing after oral drills) may not achieve transfer. Practice must match the target use context.

The explicit/implicit nuance

DeKeyser & Suzuki are careful: SAT does not claim explicit knowledge "turns into" implicit knowledge. As Paradis (2009) stresses, this never happens in the brain. Rather, explicit declarative knowledge, through proceduralisation and automatisation, can lead to knowledge that is functionally equivalent to implicit knowledge — fast, accurate, and robust — even if neurologically distinct.

Differential role by rule type

Simpler rules tend to be learned explicitly; complex and probabilistic patterns are hard to induce, comprehend, or proceduralise, and may benefit more from implicit learning (Ferman et al., 2009). SAT acknowledges a "synergy" between implicit and explicit learning — the very word Jason Anderson borrows for TATE.

Conditions for proceduralisation

Proceduralisation "cannot get started if the right conditions are not present" — it requires both the declarative knowledge AND a task setup that allows use of that knowledge (Anderson, Fincham, & Douglass, 1997). The combination of abstract rules and concrete examples is necessary to cross the "declarative threshold."

Connection to PPP

SAT is the theoretical framework Anderson draws on to justify PPP:

• Present → declarative knowledge (explicit rule presentation)
• Practice → proceduralisation (controlled exercises)
• Produce → automatisation (freer production)

This mapping makes PPP appear theoretically grounded. However, there are critical problems with this application.

Critique of the SAT → PPP Leap

From SLA research

• Interlanguage research shows acquisition follows developmental sequences that instruction cannot alter. SAT's three-stage model implies that any structure can be proceduralised at any time through sufficient practice — but developmental readiness research contradicts this
• The skill specificity finding undermines generic PPP practice: controlled gap-fills don't transfer to spontaneous conversation
• DeKeyser & Suzuki themselves note that automatisation requires "a large amount of practice" far beyond what a single PPP lesson provides

From Geoff Jordan

Jordan's critique of Anderson doesn't target SAT directly, but challenges the leap from "SAT exists as a theory" to "therefore PPP works." Even if SAT's stages are valid in principle, it doesn't follow that a synthetic syllabus presenting structures in an arbitrary order via PPP is the best way to achieve proceduralisation. SAT would predict that practice only works when the learner has the declarative knowledge to proceduralise — and developmental readiness research suggests learners can only form that declarative knowledge for structures they're ready for.

What DeKeyser actually says

DeKeyser & Suzuki are more cautious than PPP advocates:

• SAT "by no means denies a role for implicit learning"
• Complex patterns may benefit more from implicit than explicit learning
• Practice must be task-specific, not generic
• The declarative → procedural path requires the right conditions, not just presentation + drilling

References

• DeKeyser, R. & Suzuki, Y. (2025). Skill acquisition theory. In B. VanPatten, G.D. Keating, & S. Wulff (Eds.), Theories in Second Language Acquisition (4th ed., pp. 157–182). Routledge.
• Anderson, J.R. (1993, 2007). The Architecture of Cognition / How Can the Human Mind Occur in the Physical Universe?
• Paradis, M. (2009). Declarative and Procedural Determinants of Second Languages. John Benjamins.