The words hiding inside other words

Suppose you have just learned the Spanish verb construir, to build. It felt like one word, and you filed it away as one word. But look at what came in the door behind it, uninvited: construcción, a building or the act of building; constructor, a builder; constructivo, constructive; reconstruir, to rebuild. You did not really learn one word. You learned a small crowd of them — you just haven't noticed the others yet.

This is the quiet arithmetic underneath vocabulary, and it is far more generous than the flashcard model suggests. Words are not atoms. Most of them are assembled from smaller meaningful parts — a root that carries the core sense, and prefixes and suffixes that bend it into a noun, a verb, an opposite, a person who does the thing. Learn to see those parts, and every word you meet stops being a dead end and starts being a doorway.

A language is smaller than its dictionary

The most striking demonstration of this comes from a piece of counting done in 1984 by William Nagy and Richard Anderson, in a paper with the wonderfully plain title "How Many Words Are There in Printed School English?" They set out to estimate the true size of the vocabulary a school-age reader swims in — and their answer depended entirely on how you count.

If you treat every distinct written form as its own word, printed school English contains something on the order of 400,000 of them, a number so large it makes the whole enterprise of vocabulary look hopeless. But Nagy and Anderson noticed that most of those forms are relatives. Group each base word together with its inflections and its transparent derivations — build, builds, building, builder, rebuild as one household rather than five strangers — and the pile collapses to roughly 88,500 word families. The rest is arithmetic the reader gets to skip.

The vocabulary of a language is not a list to be memorised item by item. It is a smaller set of roots, endlessly recombined.

And the recombining is the norm, not the exception: Nagy and Anderson estimated that more than half of all English words are built from more than one meaningful part. This is why, in the same study, they concluded that a large share of the unfamiliar words a reader meets in text — around 60% — have meanings that can be reasonably guessed from their components. The unknown word is often just a known word wearing a hat.

The pieces are few, and they repeat

Here is the part that makes this practical rather than merely interesting: the building blocks are startlingly few. You do not need to master hundreds of affixes to unlock the doorway effect. You need a handful.

In a 1989 analysis of the words school children actually read, Thomas White, Joanne Sowell and Alice Yanagihara found that just four prefixes — un‑, re‑, in‑ (in its "not" sense, as in inaccurate), and dis‑ — accounted for about 58% of all prefixed words. Widen the net to twenty prefixes and you cover very nearly all of them — around 97%. A tiny, learnable set does almost all the work. Learn what re‑ does once, and you have quietly acquired a piece of rewrite, rebuild, reconsider, rearrange, and a thousand words you will never be formally introduced to.

The same logic runs through other languages, often more cleanly than in English. Romance languages wear their morphology on the outside: Spanish ‑ción, ‑dad, ‑mente, ‑ador attach with a regularity that turns a known verb into a predictable noun, adverb, or agent. German famously builds long words by stacking short ones, so that a learner who knows the parts can often read a word they have never seen. The affixes are the grammar of the dictionary, and grammar, unlike a word list, is finite.

What counts as knowing a word

Researchers have a tidy way of framing all this. Laurie Bauer and Paul Nation, in a much-cited 1993 paper, proposed that a word family is a base word together with its inflected forms and the derivations a learner could reasonably be expected to understand. They even graded English affixes into levels, from the most frequent and regular to the rare and irregular, precisely so that vocabulary size could be measured in families rather than in forms.

The point buried in that dry framework is a liberating one. Once you genuinely know a base word and you know the affix, the derived form is not a separate memorisation task — it is something you can generate and understand on the spot. Knowing happy and knowing what un‑ and ‑ness do gets you unhappy, happiness, and unhappiness for very little extra rent. This is the difference between learning words and learning the machine that makes words.

Does noticing the parts actually help?

It would be a nice story even if it were only a story, so it is worth asking whether teaching people to see word-parts measurably improves their vocabulary — or whether it just feels clever. The evidence points the same way the theory does. In a 2013 meta-analysis pooling dozens of studies, Amanda Goodwin and Soyeon Ahn found that morphological instruction — teaching learners to break words into roots and affixes — produced reliable gains in vocabulary knowledge, and improvements in reading comprehension besides. Later syntheses have echoed it: morphological awareness travels closely with reading comprehension across a wide span of ages.

None of these effects are miracles, and the honest summary is modest: seeing the parts helps, steadily, especially for the exact people who need it most — learners and struggling readers building a second language's vocabulary from a standing start. Which is to say, it helps you.

How to use the doorway

You do not need a linguistics course to put this to work. You need a small habit, applied at the moment you meet a new word.

• Take the word apart before you file it. When a new word arrives, look for the seams — a prefix, a familiar root, an ending that marks it as a noun or an adverb. Naming the parts once makes the whole thing stickier, and hands you its relatives.
• Collect roots, not just words. The verb is often the doorway to the noun, the adjective, and the person who does it. When you learn one, glance at the family and you will frequently find you already understand most of them.
• Learn the common affixes deliberately. A dozen prefixes and a dozen suffixes in your target language will quietly decode thousands of words. It is the highest-leverage list you can study.
• Guess before you check. Meeting an unknown word built from parts you know, try to construct its meaning first. That small act of assembly is itself a form of retrieval, and it makes the guess — right or wrong — stick harder.

There is a natural companion to all this. If word-parts are the doorways within a language, cognates are the doorways between languages — and the two stack. An English speaker who knows that ‑tion maps to Spanish ‑ción is reading morphology and cognates at the same time, unlocking whole shelves of vocabulary in a single move.

Where vocabla fits

This is why vocabla keeps words inside real sentences instead of stranding them on cards. Language in the wild is where a word's family shows up naturally — where you meet the verb one day and stumble on its noun form the next, and feel the click of recognition because you have seen the root before. Fill-in-the-blank exercises, dialogues, and definitions surround a word with the company it actually keeps, so that the pieces start to become visible on their own.

Seeing the parts is a habit anyone can build, and it changes the size of the task in front of you. The dictionary stops looking like an impossible list and starts looking like what it really is: a modest set of roots, rearranged. Learn one word well — all the way down to what it is made of — and you will keep finding, on later days, that you somehow already know its cousins. You met them the moment you met it. You just hadn't been introduced.