Decoding is the process of translating written words into spoken language by applying knowledge of letter-sound relationships. When a child encounters an unfamiliar word like "frog," they segment it into individual sounds (/f/ /r/ /o/ /g/), then blend those sounds together to produce the spoken word. Decoding is the mechanical engine of reading -- it transforms marks on a page into language that the brain can comprehend.
Start with simple CVC (consonant-vowel-consonant) words like cat, dog, and sit. Use blending drills: point to each letter, say the sound, then sweep your finger under the word while blending the sounds together. Progress to consonant blends (stop, frog), digraphs (ship, chat), and longer words. Decodable readers -- books designed to use only taught letter-sound patterns -- give children successful practice without guessing.
You already know something crucial from studying letter-sound correspondence: each letter or letter combination in English represents a sound, and those correspondences can be learned. Decoding is what you do with that knowledge — you use it to read words you've never seen before. Think of letter-sound correspondence as learning the code, and decoding as the ability to read a new message written in that code. Once you know that *f* says /f/, *r* says /r/, *o* says /ɒ/, and *g* says /g/, the word *frog* becomes readable even if you've never seen it in print. Decoding is the bridge between knowing the individual sound-spellings and using them to unlock whole words.
The mechanics of decoding involve two steps that eventually fuse into one automatic process. First, segmentation: breaking the written word into its individual phonemes by applying letter-sound knowledge. A beginning reader looks at *cat* and, letter by letter, identifies /k/, /æ/, /t/. Second, blending: holding those sounds in working memory and sliding them together into a single spoken word — *c...a...t* becomes *cat*. At first this is effortful and deliberate; experienced readers do it in milliseconds without conscious awareness. The progression from labored sounding-out to automatic word recognition is the story of early reading development, and repeated decoding practice is what drives that progression.
One of the most important insights in reading science is that decoding and comprehension are separate, separable skills — both must be strong for fluent reading. Decoding is mechanical: it converts print to sound. Comprehension is linguistic and cognitive: it extracts meaning from the spoken form that decoding produces. A child can decode *constitution* letter-by-letter and have no idea what it means; equally, a child might know the spoken word and be unable to recognize it in print. This two-skill model, sometimes called the Simple View of Reading, means struggling readers have two distinct possible failure points: decoding breakdown or comprehension breakdown. Decoding instruction targets the first, and is often the faster fix.
The reason guessing from context is an unreliable substitute for decoding is worth understanding precisely. When a child encounters *horse* and doesn't know how to decode it, guessing from a picture of a horse might work once. But reading development requires encountering thousands of words in text, many without helpful pictures, and guessing produces slow, effortful reading that never builds the automatic word recognition that fluency requires. Every successful decoding attempt — sounding out a word correctly — strengthens the brain's stored representation of that word's spelling-sound relationship. Skipping that process by guessing means the word never gets securely encoded. Decoding is not a crutch to be discarded once children improve; it is the mechanism by which words become permanently, effortlessly readable.