A frequency list helps you choose common words, but it does not group them for a particular situation. The plan is usually small: a trip in March, a job that will run in Spanish. That person needs to say “A2, food” and get a couple of dozen words that belong together, and frequency cannot do that, because it does not know that flour and recipe share a kitchen. So we sorted our 33,233-entry corpus into a grid, twelve themes across six CEFR levels. The grid is a set of decisions, and every decision has an edge where a word could have gone either way. Those edges are worth showing.
The twelve themes and what each holds
Every entry carries one theme and one level. Each of the twelve themes folds together a few finer tags from the master list: money, weather, pronouns, feelings, clothing. Each tag maps to exactly one theme, so no word sits on two shelves and none is left off.
Seven themes name the world. Food & Drink is the kitchen and the table: receta, harina, salsa, postre. Travel & Places holds places and transport, Home & Things objects, clothing, daily life and tech. Weather & Animals takes in the environment, Body & Health the senses, Numbers & Time how often something happens, People & Family the feelings. The other five carry the machinery: Greetings & Essentials for what you say first, from hello to the emergency words; Describing Things for adjectives, colours and adverbs; Everyday Verbs; Work, Money & School, which also holds education, leisure and society; and Grammar & Function Words, which gets its own section below.
Three lines were hard to draw. Money first: dinero, precio, efectivo, caro and barato went to Work, Money & School, because that is where a price gets paid; banco went to Travel & Places, because a bank is a place you walk to; cartera, the wallet, went to Home & Things, because it is an object in your pocket. Then hot and cold. Caluroso, frío, cálido and fresco sit under Weather & Animals, because we tagged them as weather, while the same fresco meaning fresh bread is a second entry under Describing Things and resfriado, the cold you catch, is under Body & Health. The theme follows the sense, so a word with two senses lives on two shelves. And then time: tiempo as weather is under Weather & Animals, and tiempo as time is tagged abstract, which puts it under Grammar & Function Words while minuto and ayer sit in Numbers & Time. That is the line we would redraw first.
Six levels, and what moves a word up one
The levels are the CEFR’s A1 to C2, and each entry gets one. The first signal is frequency. The English headword list underneath the corpus is derived from the New General Service List (NGSL 1.2, Browne, Culligan and Phillips, CC BY-SA 4.0), and the words you meet most land lowest. The labels also reflect editorial grading. In the current list, delegar, to delegate, sits at B2, while tarea, task, and colega, colleague, sit at B1. That is how these entries are labelled, not a rule that those words must be learned in that order.
Walk one theme up the ladder and both signals show. Food & Drink at A1 has café; at A2 receta and harina; at C2 umami and empalagoso, saccharine, the words of a restaurant review. Work goes from jefe and trabajo at A1 to contratar and despedir at B1, then to negociar and hito, milestone, at B2. For an everyday purchase, the beginner packs are a useful place to start.
The grade is still a judgment: the frequency list supplies the order and we drew the cut lines; the size of each rung and the reasons for it are in the piece on words per CEFR level. Where the corpus is inconsistent we would rather show it. In the English-to-Spanish direction, 177 of the 2,161 distinct word-and-translation pairs sit at two levels with the same translation, aprender as learn at A1 and again at B1. We counted every pair in every shipping manifest for that direction, Everything aggregates excluded, and kept those found under more than one level. The app treats each as its own entry, so a B1 learner meets a few A1 words again. It is the first thing we would clean.
Why function words get a home of their own
Nobody learns the theme of “the”. Yet y, pero and porque, the words for and, but and because, are among the words you hear most in any language. Spread across the other themes they would vanish: aunque, although, would land between harina and salsa because it was tagged from a food sentence, and a learner who never picked that pack would never hear it. So they have one shelf, and above A1 it is the third-largest theme in Spanish.
What sits there changes by level. At A2 it is connectors and prepositions, sin embargo, a menos que, en lugar de, and the reflexive pronouns. At B1 the shelf turns abstract: conocimiento, decisión, propósito, ley. By C2 it holds the words of argument, paradigma, dialéctica, verosimilitud. The abstract nouns are the stretch in the label; they belong to every theme and to none, which is why they share a shelf with the connectors.
Everything packs and themed packs, and why the app offers both
At each level above A1 the app offers the themed packs and one more, Everything, which is the whole rung. On the phone it is built by joining that level’s themed packs end to end, each entry keeping its theme, and it is never counted twice in our statistics. For Spanish, French and Polish, A1 is a single beginner pack, so the grid begins at A2. German, Italian and Portuguese ship A1 by theme as well.
The two shapes serve two people. The person with a plan takes a theme: A2 Food & Drink is a couple of dozen entries, and because the app leans hard towards words you have never heard, a pack that size introduces itself within days. The person who wants the level takes Everything, and at A2 in Spanish that is 625 entries per direction. At B2 and C1 the choice makes itself: those rungs hold 200 entries per language, only five of the twelve themes have a pack at all, and Everything is the sensible pick.
What the grid looks like in one language
We counted the English-to-Spanish direction, the grid as a learner meets it. Method: for every en_es pack the registry ships, we summed the entries in its manifest, skipped the whole-level A1 pack and the Everything aggregates, and grouped the rest by theme. The 1,730 entries below plus the 625 at A1 make the 2,355 of the pair’s 4,710 that run in this direction.
| Theme | Entries, A2 to C2 | Levels with a pack |
|---|---|---|
| Describing Things | 252 | A2, B1, B2, C1, C2 |
| Work, Money & School | 251 | A2, B1, B2, C2 |
| Grammar & Function Words | 248 | A2, B1, B2, C1, C2 |
| Everyday Verbs | 231 | A2, B1, B2, C1, C2 |
| People & Family | 146 | A2, B1, B2, C1, C2 |
| Greetings & Essentials | 117 | A2, B1, C2 |
| Home & Things | 108 | A2, B1, C2 |
| Body & Health | 100 | A2, B1, C2 |
| Travel & Places | 94 | A2, B1, C2 |
| Weather & Animals | 82 | A2, B1, C2 |
| Numbers & Time | 67 | A2, B1, C1 |
| Food & Drink | 34 | A2, C2 |
Three things the table says. The four largest themes hold more than half of everything above A1: adjectives, verbs, connectors and the vocabulary of work are what every level keeps adding. Food & Drink is the smallest above A1 at 34, reflecting the current pack coverage; at B1 there is no Spanish food pack at all. And the grid has a corner that is nearly empty. Numbers & Time at C1 is one entry, invariablemente, invariably. The pipeline gives every entry exactly one home, and that word’s home at C1 is Numbers & Time. A one-word pack is a strange thing to offer, and we may fold it into Everything alone.
The one pair without English
Every pair in the corpus has English on one side, except one. Polish to Italian ships at A2 as a single pack of 465 entries, for Polish speakers learning Italian. It shares nothing with the six English pairs: the Italian list was curated on its own, with Polish glosses and example sentences written for it. Its entries carry the fine tags, but it ships whole, without themed packs, so a Polish learner of Italian gets one rung with Everything on it. The grid has not reached it, and it shows what the grid costs: every themed pack at every level in both directions is curated, checked and rendered with text-to-speech before it ships.
What the grid leaves out
An entry is one word or one short headword. Pagar, to pay, is an entry. “Can I pay by card?” is what you say at the counter, and no cell of the grid can hold it, because it belongs to a situation, and to no single theme or level. The example sentence is where the grid points at this: efectivo, cash, comes with “Cash or card?” on the screen. But it is one sentence per entry, shown and never spoken on the charger, and a sentence written to illustrate a word is a different thing from a phrase chosen because you will need it.
Collocations fall through the same gap; the grid knows fecha límite, deadline, and cannot tell you which verb Spanish puts in front of it. That is why we are writing something to stand beside the grid: collections of 500 phrases each, whole sentences people say, grouped by the situations of a day and split across three lanes, survive, travel and work. They come to the blog first, as articles; packs with audio come later, with no date promised.
Using the grid with the app
The app plays one word every time the phone is plugged in, and choosing a pack is the only planning the method asks of you. Pick the rung where most words are new and a few are familiar. Then pick the shelf your plan needs: Travel & Places and Food & Drink at A2 before a trip, Work, Money & School at B1 before a job, and at B2 and C1 always Everything, the full 100 words per direction. One pack at a time is enough, because the sampler favours the words you have gone longest without hearing; one word per charge explains how the pick is made.
For the grid in numbers, the corpus statistics page has every language by level, and the per-language pages carry the themes. The methodology page covers where the list comes from and how the grades were drawn. Pick a shelf and plug in.
Common questions
What are the CEFR levels for vocabulary?
A1 to C2 is the Council of Europe’s six-level scale, from beginner to proficient. It describes what a learner can do, and whoever builds a word list maps the words onto it. In our corpus every entry carries one level: frequency puts a word low or high, and the size of each rung is our decision. In Spanish, A1 holds 1,250 entries across both directions, B2 holds 200 and C2 holds 820.
Should I learn vocabulary by topic or by frequency?
Use frequency to choose the level and topic to choose the order inside it. A frequency list tells you which words you will meet most, which is what a level is built from; a theme tells you which of those words belong together, which is what a plan needs. In Vocabcord the levels come from a frequency list and the themes sit inside each level, so you get both at once.
Where do function words like “because” go in a themed vocabulary list?
In a theme of their own. Vocabcord keeps conjunctions, prepositions, pronouns and abstract nouns in Grammar & Function Words so they can be chosen or skipped as a set; spread across the other themes they would be heard by chance or never. Above A1 it is the third-largest theme in the Spanish pair, with 248 entries in the English-to-Spanish direction.
What is the Everything pack in Vocabcord?
The whole level in one pack. On the phone the app joins every themed pack at that level end to end, each entry keeping its theme, so Everything at A2 in Spanish is 625 entries per direction. It has no download of its own and is never counted twice in the corpus statistics. Pick it when you want the whole rung in rotation; pick a theme when you have a plan.
