A Psycholinguistic Guessing Game

Learning to read is no easy feat. Reading is an unnatural act requiring years of teaching and learning to forge new pathways in the brain. There is a significant amount of research on reading instruction due to fierce disagreement about how best to teach children to read. The debate has largely been between two sides. Whole language advocates favour teaching children how to read through context. Attention to word structure is not emphasised. It is assumed that, with the right amount of effort and incidental guidance, immersion in language will ensure reading success. Those who favour the explicit instruction model, in contrast, emphasise the explicit, systematic and intensive teaching of all aspects of reading and especially the teaching of phonics to ensure mastery of the alphabetic code. Nothing is left to chance in this model of instruction. The debate between whole language and the explicit model has raged for decades and still does today.

In 1967, Kenneth Goodman published a paper titled Reading: A Psycholinguistic Guessing Game. In this paper, Goodman seeks to refute the ‘common-sense notion’ that reading is a precise process – one that requires exact, detailed and sequential identification of letters, words, spelling patterns and larger language parts (the position taken by those who adopt the explicit model). Instead, he conceptualises reading as a psycholinguistic guessing game. He writes that reading is a selective process involving the partial use of available language cues – syntactic, semantic and graphic – by the reader on the basis of their expectations while reading. During reading, decisions are made about whether to accept, reject or refine acquired information from these cues. Goodman believes it is this decision-making process that characterises efficient reading, not the reliance on text recognition, “efficient reading does not result from the precise perception and identification of all elements, but from selecting the fewest, most productive cues necessary to produce guesses which are right the first time.”

Goodman uses an analysis of student text reading to validate his claim that reading is a series of guesses informed by graphic, semantic and syntactic cues, and that reliance on precise, graphic (phonic) information will cause meaning to break down. In his analysis, the student substitutes the word the for your during reading. Goodman argues that because there is no graphic relationship between these two words, but do have the same grammatical function (both are determiners), the student must be using a syntactic cue to decode the text and that the use and subsequent error is acceptable because meaning is not lost. The ‘successful’ use of the syntactic cue is set in contrast to the unsuccessful use of graphic cueing. For example, the student makes several attempts at decoding the words philosophical and fortune without success. It was later found that the student knew the meaning of both words, information Goodman leverages to argue that if a student were to rely on semantic information during reading – if they were to make an anticipatory guess – meaning may not break down in these instances. This, therefore, validates the claim that successful reading cannot be a precise process but must incorporate all three sources of information simultaneously.

Goodman did not stop at describing reading as a balance between the 3 cues though. He believed it was poor readers who relied heavily on graphic cues and, just like in his analysis, this is why meaning often breaks down for them. So, although good readers must rely on some graphic information, they only sample just enough to confirm an informed guess from syntactic and semantic information, “as the child develops reading skill and speed, he uses increasingly fewer graphic cues…” therefore, “skill in reading involves not greater precision, but more accurate first guesses.”

Goodman’s assertion that good reading is not precise but relies on a series of guesses has been completely crushed in the scientific research literature. For many years leading up to Goodman’s paper, rhetoric and intuition reigned because decisive evidence was hard to obtain. However, the years following consisted of research with several types of findings which converged on the same conclusion: phonological information is an essential element in skilled reading and impairments in the use of phonological information are typical of poor readers. No research has argued that attention to word structure is all reading is; rather, it is an essential part of reading and one crucial characteristic of skilled reading. Good readers do not rely on a series of guesses; they rely on precise and detailed attention to letters and words. Guessing informed by syntactic and semantic cues is used by poor readers to compensate for their poor decoding ability.

Charles Perfetti led some breakthrough research that showed it is indeed poor readers, not good readers, who rely heavily on context to compensate for their poor decoding ability. Perfetti studied good and poor readers’ ability to read single words categorised as common, less common or non-words. Both good and poor readers read the common words equally well, but the poor readers did much worse reading the less common and non-words. Later studies yielded the same results when the words were embedded in context (within sentences). Good readers were faster and more accurate overall, as expected. Poor readers relied much more on context to read less common words and were far less accurate. The message from these studies is clear: good readers rely less on context and read much more accurately; poor readers rely more on context and read less accurately. These studies’ findings have been confirmed many times and are well known amongst reading researchers, yet they are still not aligned with how students are taught to read.

Goodman’s psycholinguistic guessing game was simply a bad idea invalidated by scientific research. It should be history – and in reading-related fields, it is – but bad ideas die hard in education. Although most educators have not heard of Kenneth Goodman or the PGG, they are sure to be very familiar with the semantic, syntactic and graphic cues associated with it. Yes, all educators are familiar with the 3 cueing system with its 3 circles intersecting at the centre to make meaning. It is considered the ideal model for teaching reading, incorporating a balance between all three cues. The model (and how it is used) is incredibly similar to how Goodman describes his PGG and is, therefore, predicated on notions of reading that have long been discredited, just like the PGG.

How this model has damaged the teaching of reading cannot be overstated. If the questions and prompts listed in the above table aren’t enough to convince you, please take the time to consider the texts students are introduced to when they initially learn to read. They are designed to accommodate this model of reading – one where students are encouraged to guess using cues long demonstrated to be characteristic of weak readers and not good readers. We are literally encouraging students to use the compensatory strategies common amongst poor readers from day 1 of kindergarten. There are many students who will overcome this obstacle, but some will really struggle. Having them engage in the psycholinguistic guessing game will only make things worse, not better.
Why the 3 cues written about by Goodman in 1967 are so popular 50 years later is a topic worth discussing. We know that this theory of reading does not align with common understandings of skilled reading and how one learns to read, so why does it persist? Why did I, like so many, read about it in my assigned textbook at university? Why do teachers hold onto it like a childhood toy? I think teacher educators have a lot of explaining to do, and that is certainly the position taken by Mark Seidenberg. In his book, he has a chapter titled The Two Cultures of Science and Education. In this chapter, he talks about how schools of education so readily preference their personal ideologies – often quite romantic – over the scientific consensus on how we come to read. The 3 cueing system is a microcosm of a much bigger cultural problem, one where ideology reigns and evidence does not:


“The 3-cueing approach is a microcosm of the culture of education. It didn’t develop because teachers lack integrity, commitment, motivation or intelligence. It developed because they were poorly trained and advised. They didn’t know the relevant science or had been convinced it was irrelevant. Lacking this foundation, no such group could have discovered how reading works and how children learn.” (Seidenberg, 2017, p.304)

When will education change tack? Reading is most definitely not a guessing game. It is time to stop the guessing. It is time to embrace scientific evidence. Surely, after 50 years, it is time to move on.

Goodman’s Pyscholinguistic Guessing Game

Seidenberg’s Language at the Speed of Sight

Perfetti’s research (1980)

1 thought on “A Psycholinguistic Guessing Game”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s