Monday, 22 September 2014

If learning to read is a “linguistic task”, what’s wrong with Whole Language?

Although it is some years since I have worked clinically, my initial qualification was in Speech Language Pathology and I have always had a particular interest in, and passion for language – both spoken and written. I love spending time with infants and toddlers and observing their amazing emergent grasp of the linguistic code – surely one of the greatest wonders of being human. My research activities in recent years have had a strong focus on “vulnerable” learners and I’ve had the enormous privilege of spending a lot of time in primary schools, interacting with teachers about (in particular) the teaching of reading in the early years.

A large focus of the presentations I give to teachers is explaining the linguistic basis of learning to read. This entails discussion of the fact that unlike learning to talk and understand, learning to read and write is biologically “unnatural” – i.e. humans are pre-programmed in an evolutionary sense for the former, but not the latter. This concept is often quite a revelation for teachers and I see many “ah hah!” moments in the course of these discussions.
I often use an analogy of a child having to “cross a bridge” in the early years of school – from talking and listening on one side, to reading and writing on the other. If I’m feeling particularly prosaic on a given day, I might mention the crocodile-infested waters underneath….

While some fortunate children will metaphorically “skip” across the bridge and suddenly find themselves on the other side without having had much awareness that they were even on a bridge, many will find the experience one that requires conscious focus, guidance and support from adults, and carefully-paced milestones. Complex genetic and social factors no doubt account for such variability, but classrooms need to cater to all children, not just the very able.
Proponents of Whole Language-based approaches to the teaching of reading argue that reading is as natural as talking, and so simply requires immersion in text-rich experiences, in order for children to find themselves on the far side of the bridge.

This argument fails a number of reality checks however, the first of which is that children’s acquisition of oral language, while strongly “pre-programmed” in an evolutionary sense, still requires enormous environmental input and interaction, in the form of exposure, simplification, repetition, imitation, expansions, recasts, and so on. There is something of a “dose-response” relationship between early language exposure and verbal skills in early childhood, such that this is a phenomenon that sits on a social gradient (see references below).  
Secondly, decades of cognitive psychology research tells us that the acquisition of skilled reading draws heavily on the child’s psycholinguistic knowledge about how their language works. Now of course we can’t sit down and interview a four-year old about her psycholinguistic knowledge, but we can extrapolate important information from spontaneous language samples and from the 1:1 administration of formal language measures.
By “psycholinguistic knowledge”, we are referring to the child’s grasp of grammatical rules (e.g. in English, the fact that in order to make a noun plural, we generally add an “s”), knowledge of narrative structure (the logical and sequential way in which experiences are organised and presented for the benefit of a listener), vocabulary use and comprehension, and knowledge of grammatical conventions to link words and express increasingly complex (e.g. embedded) ideas. Knowledge of the sound system (phonology / phonemic awareness) is critical, as it is the means by which a child can segment and blend sounds to manipulate and integrate word components. Finally, knowledge of pragmatics teaches the child the subtle and sometimes confusing social conventions around how language is used in a variety of communicative contexts.  We typically learn many of these rules through the process of breaking them, and then receiving some kind of corrective input from a nearby adult, such as a parent or teacher.

Bear in mind too that all of these psycholinguistic milestones need to be mastered in both the receptive (comprehension) and expressive (spoken) modalities. Further, children need to master the understanding and use of literal language as well as figurative, or idiomatic language. Everyday communication would be quite boring if we only ever said exactly what we mean. Instead, we make our exchanges more interesting and entertaining (but often more confusing!) by using linguistic devices such as metaphor, sarcasm, and humour.
As children's grasp of oral language develops, they acquire the ability to think and talk about language and how it works, i.e., metalinguisitic abilities, and these can then be drawn on in the early years classroom and in everyday life, e.g. to explain why a joke is funny. 

Oral language, then, is one of those deceptively complex processes that by virtue of its pervasive presence in our lives can appear simple, or even inconsequential, as a developmental achievement.  
It is neither.
So when educators argue the virtues of Whole Language as a reading instruction approach, they are conveniently lumping together a raft of linguistic microskills that the child is (a) still acquiring and refining and (b) needs to draw on in order to cross the bridge from talking and listening on this side, to reading and writing on the other. To the untrained eye, however these skills are not observable or accessible.
The fact that “Whole Language” has the word “language” in its name is both unfortunate and misleading, because learning to read IS a linguistic task, but Whole Language does not deliver, unpack, decipher and value-add the linguistic goods that all children need to acquire in order to learn to read. The fact that some (fortunate) children will learn to read in Whole Language classrooms simply attests to the enormous variability that teachers in early years settings need to address.
My driving motivation is to provide all children with access to the life-changing opportunities that await their arrival on the other side of the bridge. I'm sure this is the driving motivation of all who are invested in early years education, however examination of the research evidence shows uneven results.
That Whole Language is an approach that fosters passion about children's literature is laudable. However, direct instruction by teachers who understand the psycholinguistic properties of the language of instruction is key to shepherding more children to the other side of the reading bridge, safe from those crocodile-jaws of early academic failure.

Hart, B. & Risley, T. (1995). Meaningful Differences in Everyday Parenting and Intellectual Development in Young American Children. Baltimore: Brookes.
Hoff, E. (2003). The specificity of environmental influence: Socioeconomic status affects early vocabulary development via maternal speech. Child Development (74), 1368-1378.  
Locke, A., Ginsborg, J. & Peers, I. (2002). Development and disadvantage: Implications for the early years and beyond. International Journal of Language and Communication Disorders, 37, 3-15.
Weisleder, A. & Fernald, A. (2013). Talking to children matters: Early language experience strengthens processing and builds vocabulary. Psychological Science, 24(11), 2143-2152.

(C) Pamela Snow 2014

Monday, 8 September 2014

Reading Recovery and Cassandra's Curse

Last week, I attended an excellent, though quite depressing seminar organised by Learning Disabilities Australia, entitled Reading Recovery and the Failure of the New Zealand National Literacy Strategy, presented by Professors Bill Tunmer and James Chapman (both of Massey University). A summary of the content of the presentation can be found in this paper.

The seminar was very well attended, with senior policy folk from the state and Catholic sectors present, as well as a number of education and speech pathology academics and practitioners, teachers, tutors, and other interested parties. The questions and comments from the floor were many and I think people would have stayed on for hours continuing the discussion if that had been possible. 

So why did I find it depressing? The first depressing aspect was the 
established evidence on Reading Recovery's failure to deliver (as per PIRLS data) on the stated aims of its developer, the late Dame Marie Clay (1993), to 

" children acquire efficient patterns of learning to enable them to work at the average level of their classmates and to continue to progress satisfactorily in their own school's instructional program. When recognized training accompanies the use of procedures contained in this book, success rates are consistently high and surprising."

The second depressing aspect was the fact that policy makers (and as a consequence many, though by no means all, teachers) have been slow to acknowledge and act on the irrefutable data that Reading Recovery is not delivering on its promises. That it is in particular, leaving behind children who come from low socio-economic status families and schools is especially disappointing and unforgivable. 

No-one would disagree with the importance of helping struggling learners to catch-up, and the earlier the better. That's where the consensus in this debate ends, however, and the so-called "reading wars" saga rears its ugly head. I don't intend to deconstruct this debate here (plenty of metaphorical ink has already been spilled on that one), but it is important to note that Reading Recovery is a "child" of the Whole Language movement. In essence, this approach to teaching reading, espoused in the late 1960s and 1970s by Kenneth Goodman stemmed from the attractive (though unsubstantiated) claim that children can acquire literacy skills in much the same way that they acquire oral language skills - through experiential immersion. While this idea may have had strong face validity and appeal to the educational zeitgeist, it lacked one critical ingredient - an empirically derived theoretical basis. In fact, we now know (and probably knew back then) that unlike learning to speak and understand, learning to read and write are biologically "unnatural" acts (Gough & Hillinger, 1980) requiring sustained specific instruction. As Melbourne Speech Pathologist Alison Clarke explains in 
this engaging YouTube clip phonics-based reading instruction was suddenly deemed "old fashioned" and was pretty much universally abandoned in Australian schools (and teacher-training courses) in the 1970s, in favour of Whole Language-based approaches. 

Many aspects of Reading Recovery have long perplexed me, most notably the question of what is being "recovered" for children who had not acquired the requisite skills in the first place? Secondly, I don't understand why any education system would adopt a reading instruction approach (Whole Language) that assumes a 15-20% failure rate (and need for expensive and intensive intervention) by the end of the first year of school ("failure" here referring to the lowest performing students in the class - those whom Clay argued could be brought up to the norm via Reading Recovery). Why not simply employ effective, evidence-based teaching approaches for all, and move the whole curve to the right?

Imagine a hospital administrator deciding that operating theatre infection control guidelines were "old-fashioned" and unwieldy, and so proclaiming that henceforth, we'll dispense with strict hand-washing protocols, letting practitioners use their own judgement on this, and rely instead on people's natural immunity. As a fall-back, anyway, we have postoperative antibiotics for the tail of the surgical curve who (inevitably) happen to develop a serious infection. Of course those in that tail will already be disadvantaged in other ways, e.g., being older, or having multiple comorbidities, making their immune systems compromised and their benefit from antibiotics equivocal at best, and non-existent at worst. Let's not even try to calculate the cost to the system (and society) of increased length of stay and lost social and economic productivity. 

Of course such a situation would not occur and most would read that scenario and consider it laughable. But why are health and education so different with respect to the way in which evidence is derived, critiqued, and applied?

One key factor might be that when hospitals adopt systems that don't work, patients die, and people can see that patients die. When schools adopt systems that don't work, no-one dies and practitioners are largely shielded from the detrimental effects of their practices, however pernicious these effects may be over the longer term.

However just as we can't hold individual doctors accountable for policies enacted in hospitals, we can't hold individual teachers to account for policies enacted by their schools or education sectors. We must, however, hold policy makers and those who are responsible for teacher training to account to apply evidence ethically. The Whole Language-Reading Recovery grip on teacher education and early years practice is doing a huge disservice to the most vulnerable and socially disadvantaged learners. These are the very learners who need school to 
accelerate their progress relative to more advantaged peers. The evidence presented by Tunmer and Chapman shows unequivocally that such acceleration is not occurring in NZ, and there is no evidence (of which I am aware) to indicate that it occurs here either. 

I teach medical students enrolled in an undergraduate MBBS. Putting aside for a moment differences in university entrance requirements for medicine Vs teaching, a major difference I see between these students and those in education (whom I have also taught) is their capacity and willingness to question and challenge evidence that is presented to them. In medicine, we know we would be producing graduates who are frankly dangerous to the community if they do not exit university as astute life-long consumers of new research and of edicts from "on high" (e.g. hospital administrators) that they should adopt or abandon a particular practice. Student doctors are taught to "argue their corner" and to do so using the language of evidence, not anecdote. Treating a medical condition is a matter of science (blended with positive personal qualities such as warmth and respect), not a matter of personal ideology (with a few notable exceptions, such as termination of pregnancy, where guidelines exist for how doctors should deal with their personal biases so as not to disadvantage their patient).

In the case of reading instruction however, we seem to have a terrain that has elements of religious boundary-setting around what teachers (and their educators) can and cannot be challenged over. This is ethically indefensible and if allowed to continue will simply perpetuate the widening gap between the "haves" and "have nots" in the early years classroom and beyond. There is nothing progressive, socially or educationally, about standing by and being complicit in maintaining such an unfair status quo.

Academics who have interrogated the evidence on Reading Recovery and found it to be wanting seem to have been afflicted by Cassandra's Curse - the ability to predict the future alongside the sure knowledge that they will be ignored.

References and further reading

Buckingham, J., Wheldall, K. & Beaman-Wheldall, R. (2013).Why Jaydon can't read: The triumph of ideology over evidence in teaching reading. Policy29(3), 21-32. 

Carter, M. & Wheldall, K. (2008). Why can't a teacher be more like a scientist? Science, pseudoscience and the art of teaching. Australasian Journal of Special Education, 32:1, 5-21.

Clay, M. M. (1993). Reading Recovery: A Guidebook for Teachers in Training.
Portsmouth, NH: Heinemann.

Gough, P. B., & Hillinger, M. L. (1980). Learning to read: An unnatural act. Bulletin of the Orton Society, 30, 179-196.

Learning Disabilities Australia 

Reynolds, M. & Wheldall, K. (2007). Reading Recovery 20 years down the track: Looking forward, looking back. International Journal of Disability, Development and Education, 54, 199-223.

Tunmer, W.E., Chapman, J.W., Greaney, K.T., Prochnow, J.E., & Arrow, A.W. (2013). Why the New Zealand national literacy strategy has failed and what can be done about it: Evidence from the progress in international reading literacy study (PIRLS) and Reading Recovery monitoring reports. Australian Journal of Learning Disabilities, 18(2), 139-180.

(C) Pamela Snow 2014