When children very first learn that symbols can be used to represent ideas or spoken words, they begin with the assumption that the whole symbol represents the whole word. Because it is so difficult to memorize a complete, complex shape like a word, children adopt strategies of seeking out salient features from a word, and using those features to identify the word. This is why children can recognize the word “McDonalds” when the M is represented by the golden arches, but fail to recognize it when it is printed for them, and this is why children often mistake words like LOOK for BOOK or WAG for DOG. Children almost instinctively attempt to memorize words as wholes when they can, and they seek out distinctive features of words.
Philip Gough conducted a brief study that revealed these tendencies in children. He asked children to learn some made-up vocabulary words, which he presented on flash-cards. In the corner, on one card, he deliberately placed a thumb-print. Children were very quick to learn the word which was on the card with the thumb-print, but after the children learned that word, they typically did not recognize it when the thumb-print was removed. Further, when the thumb-print was moved to a different card with a different word, children tended to erroneously call the name of the word that originally accompanied the thumb-print. Even more revealing, however, when Gough presented a card containing only a thumb-print, and no word, children still tended to call the name of the word they had originally associated with the thumb-print.
Clearly this strategy for learning new words is maladaptive. Children memorize a word that is highly dependent upon context, and because most words share many visual features with many other words, children who attempt to memorize words as wholes tend to confuse words. Moreover, there is a limit as to how many words children can memorize — while most competent readers have a reading vocabulary of around 50,000 words (see V is for Vocabulary), children who memorize words as wholes are only capable of learning a maximum of about 5,000 words in isolation.
In order to become competent readers with reading vocabularies in the 50,000 to 75,000 range, children need to learn to decode words rather than simply memorizing them. Decoding words is much more generative and flexible and requires much less attention and memory. Children who can decode words are able to break down new, unfamiliar words, and arrive at a phonological code that they can communicate with others (i.e. a child can sound out an unfamiliar word, and, if necessary, ask others what that word means).
One important goal then in teaching children to read is to encourage them to abandon their natural tendency to memorize words as wholes, or to memorize salient features of words, and instead to learn to break words apart, examine the letters and chunks within the words, and decode them.
How, then, are we to explain the time and effort spent teaching children to memorize words? An often stated goal of many reading teachers, reading programs, and even state standards documents, is that the teacher will enhance the child’s repertoire of “sight words.”
The concept of sight words has foundations in the “Look-Say” approach to reading instruction — the idea was to teach children to simply memorize the most common words in written English on the assumption that memorizing the most common words in the language would give the child a leg up when attempting to read connected text. A child’s natural tendency to memorize the whole word, or to memorize some salient feature of the word, was encouraged by teachers, and to facilitate the memorization of the words, children were presented with text that was composed almost entirely of words from the popular sight word lists. Children were able to read those texts, but usually had difficulty reading more authentic text which was not primarily composed of sight words.
The term “sight word” is still with us, and the sight word lists that were created before World War II, such as the Dolch list, are still very popularly used. However, some people have reinterpreted the definition of a sight word. Whereas a sight word once universally referred to a word which the child had memorized as a whole (without learning to decode it), now some have redefined the term to mean something different.
Some use the term “sight word” to refer to words which do not adhere well to the principles taught in phonics lessons (e.g. WAS, THE, ONE, OF, SHOE, SAID), and which must, it is therefore claimed, be memorized. These words have traditionally been called “irregular” words, or “exception” words, but some are also applying the term “sight word” to words in this category.
Some use the term “sight word” to refer to words which have been encountered so frequently that a reader no longer needs to laboriously sound them out. The first time a child reads the word YELLOW, the child may struggle and have some difficulty. Gradually, the child becomes more and more familiar with the word, and eventually, the child is able to read the word without hesitation or conscious thought. At the end of this evolution, according to this perspective, the word becomes one of that child’s “sight words.”
Neither of these applications of the term seems appropriate. Words that can not be directly sounded-out already have a designation, they are called exception words or irregular words, and even these exception words are not memorized as wholes — most of the letters in exception words are “regular” and children still benefit from processing these words at the letter level, chunking the words when necessary, and applying knowledge of letter-sound relationships.
Likewise, the notion that sight words are words which are processed so automatically that no conscious thought is required also seems specious. By this definition, pseudowords like BIP and FANK are sight words for most skilled readers because, even though they have never encountered those words before, skilled readers able to process them automatically without concerted effort.
The term “sight word” has a clear definition, and adopting that term for other concepts only serves to confuse the issues. If a child has learned to recognize a word without learning to decode the word, then that word is a “sight word” for that child. When a teacher encourages a child to memorize more words by sight, that teacher is delaying the inevitable — eventually, in order to become a good reader, that child must begin processing words at the letter level.
There is no clear empirical evidence that teaching very young children to memorize a few basic and common words is harmful — for very young children, this approach may actually help to build a foundation and familiarity with text. However, it seems clear that teaching children in the first and second grades to memorize words only detracts from one of the primary goals of reading instruction — as early as possible, children need to learn to attend to the letters within the words, and to decode the words, and children need to become so proficient at this skill that words are decoded rapidly and effortlessly.