Spaced Repetition for Portuguese: The Complete System
Spaced repetition is the most research-validated vocabulary learning method available. But most learners implement it wrong — optimising the review schedule while ignoring the encoding quality that determines how many reviews each word actually needs.
What Is Spaced Repetition?
Spaced repetition is a learning technique that schedules review of information at increasing intervals, timed to occur just before the point of forgetting. The technique is based on two closely related discoveries in cognitive psychology.
The Ebbinghaus Forgetting Curve
In 1885, German psychologist Hermann Ebbinghaus conducted the first systematic study of memory and forgetting. He memorised thousands of nonsense syllables and tracked how quickly he forgot them. The result was the forgetting curve: a steep exponential decay that shows the average person retains only about 40% of new information after 24 hours, dropping to roughly 20% after a week without review.
Ebbinghaus also discovered the spacing effect: if you review information at intervals rather than massing review all in one session, retention improves dramatically for the same total study time. Studying a word for five minutes spread across five days produces better long-term retention than studying the same word for five minutes in a single session.
The Spacing Effect
Each time you successfully recall a piece of information, the next forgetting curve resets at a shallower slope — meaning the same information takes longer to forget. Spaced repetition exploits this by scheduling each review at the optimal moment before forgetting, so every review session is maximally efficient.
From Ebbinghaus to SM-2 and FSRS
Piotr Wożniak formalized the spacing effect into a computable algorithm in the late 1980s, producing the SM-2 algorithm that underlies Anki and many other SRS tools. Modern research has refined this further: the Free Spaced Repetition Scheduler (FSRS), published in 2022, uses a more accurate model of memory decay that accounts for individual variation in memory strength, producing better interval predictions than SM-2 — particularly for difficult items that learners keep failing.
Why Portuguese Specifically Benefits from SRS
Spaced repetition benefits any language learning. But Portuguese has several features that make a systematic SRS approach particularly valuable.
Large active vocabulary requirement
Conversational fluency in Portuguese requires roughly 2,000 to 3,000 words. Reading comprehension of newspapers requires 5,000 to 8,000. Without a systematic acquisition system, reaching these targets is extremely slow. SRS makes it tractable by ensuring every word you study moves reliably toward long-term retention rather than cycling through short-term memory.
Complex verb conjugation families
Portuguese has six major verb conjugation groups and significant irregularity in the most common verbs. The conjugation forms you need to recognise passively (in listening and reading) and produce actively (in speaking and writing) are largely separate memorisation tasks. SRS handles both efficiently, especially for irregular forms that resist rule-based learning.
Gender, diminutives, and augmentatives
Each Portuguese noun has a gender (masculine or feminine) that affects article and adjective agreement throughout a sentence. Gender assignment is often unpredictable from the noun form. SRS — particularly when gender is encoded into the flashcard design — is the most reliable way to learn noun genders correctly and retain them.
False cognates with English
Portuguese has hundreds of words that look like English words but mean something different. 'Borracha' looks like 'baroque' but means 'rubber'. 'Polvo' sounds vaguely like 'pulp' but means 'octopus'. These false cognates are interference-prone and require deliberate, spaced overlearning to encode correctly.
The Science: Optimal Review Intervals
SRS algorithms calculate the optimal interval before each review. The key insight is that the interval should be as long as possible without exceeding the point where forgetting becomes likely. A review conducted too early wastes time (the memory was stable and didn't need reinforcement). A review conducted too late means the word has already been forgotten and must be re-encoded from scratch.
| Review number | Typical interval (new word) | Memory state |
|---|---|---|
| Review 1 | 1 day | Short-term — easily disrupted |
| Review 2 | 3 to 4 days | Consolidating — still fragile |
| Review 3 | 7 to 10 days | Medium-term — more stable |
| Review 4 | 2 to 3 weeks | Semi-durable — most learners stop here |
| Review 5+ | 1 to 3 months | Long-term — approaching permanent |
The intervals above assume standard flashcard encoding — a shallow initial trace. Words encoded in a memory palace start with a much deeper trace, which means the first interval can be longer and the total number of reviews to reach permanence is lower. This is the compounding advantage of combining encoding quality with scheduling precision.
SRS Alone vs. SRS + Memory Palaces
Spaced repetition and memory palaces solve different parts of the vocabulary learning problem. SRS optimises when you review. Memory palaces optimise how deeply a word is encoded on first encounter. The two are not alternatives — they are complements.
A word studied with a flashcard arrives in your SRS system with a shallow encoding. It will need to be reviewed frequently — often every day or two for weeks before the interval extends to a week. A word encoded in a memory palace arrives in your SRS system with a deep encoding: spatially anchored, narratively scaffolded, and emotionally engaged. It is already stable enough to survive a week-long interval from the first review, requiring far fewer total reviews to consolidate.
SRS alone
- Optimises review timing
- Depends on shallow flashcard encoding
- High review load — 5 to 8 reviews to consolidate
- Short intervals early on (daily reviews)
- Effective but time-intensive
SRS + memory palaces
- Optimises review timing AND encoding depth
- Deep initial encoding via spatial narrative
- Lower review load — 2 to 4 reviews to consolidate
- Longer intervals possible from the start
- Faster to permanence, more robust under time gaps
For a detailed breakdown of how these two systems interact, see Spaced Repetition + Memory Palaces: The Ultimate Vocabulary Stack.
How Loci's SRS Engine Works
Loci uses an FSRS-inspired scheduling algorithm that tracks each word's individual memory stability and retrieval difficulty, and uses these to predict the optimal next review interval. The algorithm produces more accurate intervals than SM-2 — particularly for words that you consistently find difficult, where SM-2 tends to over-penalise and schedule reviews too frequently.
Crucially, the scheduling works on top of memory palace encoding, not instead of it. When you first encounter a word in Loci, you walk its memory palace scene. The scene creates a deep initial encoding. The scheduler then calculates the optimal first review interval based on the strength of that encoding — typically longer than the first interval an Anki deck would assign, because the encoding is stronger.
Loci Review Flow
Walk
You encounter the word in its memory palace scene — keyword, narrative, native audio. Deep spatial encoding occurs.
Recall
At the scheduled review, you are tested with active recall — you must produce the word before seeing the answer. The act of retrieval strengthens the trace.
Rate
You rate your confidence (again / hard / good / easy). The algorithm uses this to update the memory stability estimate for this specific word.
Schedule
The next review interval is calculated from the updated stability estimate. Stable words are scheduled further out; unstable words are reviewed sooner.
Sample 30-Day Portuguese Vocabulary Plan Using SRS
This plan assumes 20 to 30 minutes of study per day and a target of 15 new words per day. It is structured around Loci's thematic palaces, which group semantically related words into the same location for compounding spatial and semantic context.
Week 1 (Days 1–7)
~105 wordsFocus: Foundation vocabulary
- – Core function words (pronouns, articles, prepositions, common conjunctions)
- – Numbers 1–100
- – Days, months, time expressions
- – Common greetings and social phrases
Build the palace skeleton. Review pile is still small — use extra time to practice mental walks.
Week 2 (Days 8–14)
~105 new + ~80 reviewsFocus: Daily life vocabulary
- – Food and drink (kitchen palace)
- – Home objects and rooms (home palace)
- – Body parts and health basics
- – Common adjectives (big, small, good, bad, fast, slow...)
Reviews begin to accumulate from Week 1. Total daily session is now 25–35 minutes.
Week 3 (Days 15–21)
~105 new + ~150 reviewsFocus: Movement and action
- – High-frequency verbs: ser, estar, ter, fazer, ir, querer, poder, saber
- – Transport and travel vocabulary (transport palace)
- – Shopping and money (market palace)
- – Common adverbs and question words
Verb conjugation forms — especially irregular preterite — should be added as separate SRS cards.
Week 4 (Days 22–30)
~135 new + ~200 reviewsFocus: Social and emotional vocabulary
- – Emotions and mental states
- – Relationship and social vocabulary
- – Work and professional basics
- – Weather and nature
By the end of Week 4 you have ~450 words in your SRS system. Week 1 words are now at 2–3 week intervals.
Common SRS Mistakes (And How to Avoid Them)
SRS is powerful but its effectiveness depends entirely on consistent, disciplined use. These are the mistakes that most commonly derail learners.
Adding too many new cards at once
New cards create a review debt that compounds over days. Adding 50 cards in one enthusiastic session creates a wall of reviews 3 to 5 days later. Cap new cards at 15 to 20 per day until you have several weeks of data on your sustainable review load.
Ignoring difficult items
SRS systems surface difficult items more frequently precisely because they are difficult. Skipping, burying, or marking them as 'easy' to avoid them produces a vocabulary full of holes — words you believe you know but will fail under time pressure or conversation.
Letting a review backlog accumulate
A one-day gap in reviews creates a pile. A one-week gap creates an avalanche. If you fall behind, reduce new cards to zero and clear the backlog before adding more. If the backlog is unmanageable, use the 'bury' function on all cards except overdue ones and work through them systematically.
Using passive re-reading instead of active recall
The entire value of SRS depends on active retrieval — attempting to recall the answer before seeing it. Flipping cards immediately and scanning the answer is rote repetition with a scheduler attached. Force yourself to produce the word or its meaning before revealing the answer, every time.
For a comparison of Loci versus a pure SRS tool and how the encoding difference affects your review load, see Loci vs. Anki. For vocabulary retention techniques that complement SRS, see How to Remember Vocabulary: 7 Science-Backed Techniques.
Frequently Asked Questions
Common questions about using spaced repetition for Portuguese vocabulary.
How many new Portuguese words should I add per day with SRS?
For most learners, 10 to 20 new words per day is the sustainable sweet spot. Below 10 and progress is slow; above 30 and the review load from previous days accumulates faster than you can clear it. The specific number depends on your study time: each new card adds roughly 2 to 4 review repetitions over the following weeks. At 20 new words per day, expect a daily review pile of 80 to 120 cards within 2 to 3 weeks. If reviews start to pile up, reduce new cards until the backlog clears before adding more.
Is Anki or Loci better for learning Portuguese with SRS?
Anki is a flexible, powerful SRS tool with a large community deck library for Portuguese. Its limitation is that it only handles the scheduling side of vocabulary learning — the review timing — but relies entirely on you (or a deck creator) to provide meaningful encoding. Most Anki decks present words as isolated text pairs, which produces shallow encoding and high review loads. Loci combines SRS scheduling with memory palace encoding: every word comes with a pre-built vivid scene that creates deep initial encoding, so each word needs fewer reviews to consolidate. For learners who want maximum retention with minimum review time, Loci's approach is more efficient. For learners who want maximum flexibility and a DIY system, Anki remains excellent.
Do I need to learn grammar separately, or can SRS handle conjugations too?
SRS handles vocabulary and memorisable grammar facts (irregular conjugations, exception rules, preposition choices) very well. It is less suited to internalising grammar patterns that you need to apply productively in real-time — for example, knowing when to use the subjunctive versus the indicative in novel sentences. The practical approach is to use SRS for vocabulary and memorisable grammar items, and use comprehensible input (listening and reading at the right level) to internalise productive grammar patterns through exposure. The two methods are complementary, not competing.
How long will it take to reach 2,000 words in Portuguese using SRS?
At 15 new words per day with consistent daily review, you would reach 2,000 words in approximately 4.5 to 5 months. In practice, learners typically take 6 to 9 months because of missed days, pace adjustments, and time spent on grammar and listening alongside vocabulary. The 2,000-word milestone is significant: research suggests this is roughly the vocabulary needed for comfortable comprehension of everyday spoken and written Portuguese. The first 500 words take about 5 to 6 weeks; the path from 500 to 2,000 is longer per word because the words are less frequent and appear less often in natural language to reinforce learning.
Further Reading
Loci Language App
SRS scheduling + memory palace encoding, built in.
Loci handles both sides of Portuguese vocabulary retention: deep spatial encoding on first encounter, and FSRS-inspired scheduling for every subsequent review. No Anki deck setup. No scene construction. Just 15 minutes a day.