The Futility of the Spelling Grind Stephen Krashen, Ph.D. Emeritus Professor of Education University of Southern California ----- Original Message ----- From: Stephen Krashen Sent: Thursday, November 21, 2002 12:54 AM Subject: The futility of the spelling grind Sent to Education Week, November 20, 2002 In Ed Week's report on a to-be-published study ("Studies back lessons in writing, spelling," November 20), Steve Graham notes that diverting attention to spelling while writing "disrupts the planning process." His cure is to spend more time on direct spelling instruction. There is another possibility: Advise writers to delay focusing on correct spelling until their ideas are firmly in place, while, at the same time, building up spelling competence through massive reading. There is good support for each of these suggestions. A number of studies show that good writers delay editing concerns until the final draft, and "premature editing" has been shown to be a predictor of the frequency of writing blocks. Mike Rose found this was the case for writers in English, and Sy-ying Lee of National Taipei University has shown that premature concern with form and editing relates to writing blocks for writers in Chinese, and for writers in English as a foreign language. There is also very good evidence that direct instruction in spelling has limited effects. The research begins with J.M. Rice's study, "The futility of the spelling grind," published in 1897, showing no relationship between the amount of time devoted to spelling and spelling achievement, when measured on tests involving words in sentences and compositions. It also includes Oliver Cornman's study, published in 1902, showing that dropping formal spelling instruction had no effect on spelling accuracy, whether measured in isolation or in compositions. In a 1991 paper, Howard White and I re-analyzed Rice's and Cornman's data using modern statistics, and confirmed their results. Since then, it has been shown that children acquiring their first language and adult second language acquirers can improve in spelling without instruction. In 1982, for example, Kenneth and Yetta Goodman reported that one of their children learned to read and spell before coming to school, without any formal instruction, and in 1991 Marlene Haggan showed that fourth year Arabic-speaking English majors at the University of Kuwait made fewer spelling errors in their writing than first-year students, even though little emphasis was put on explicit teaching of spelling in the curriculum. In 1977, Donald Hamill, Stephen Larsen and Gayle McNutt reported that children who had spelling instruction spelled better than uninstructed students in grades 3 and 4, but the differences disappeared by grades 4 and 5. This suggests that spelling instruction, when it works, only succeeds in helping children learn to spell words that they would have learned to spell on their own anyway. Additional evidence comes from Sandra Wilde's work: Wilde estimated that each spelling word learned through direct instruction takes about 20 minutes of instructional time. Given the huge number of words, this result strongly suggests that instruction cannot do the job. A focus on spelling rules is equally hopeless. W. Cook, back in 1912, tested high school and college students who had just completed a semester of intensive study of spelling rules. There was no difference in spelling accuracy among those who said they knew the rules and used them, those who said they knew the rules and did not use them, and those who said did not know the rules. He also found that even though the students had just studied the rules, many could not recall them. When asked to state the rules, they typically gave versions much simpler than the complex rules they had been taught. The most likely candidate for building spelling competence is reading. This conjecture is supported by studies showing that each time readers read a passage containing words they cannot spell, they make a small amount of progress in acquiring the correct spelling, as well as studies showing that our spelling gets worse when we read misspelled words. Also, some studies show positive correlations between spelling competence and the amount of reading done. I am fully aware that many people who are well-read are not perfect spellers. My claim is that reading will make you a very good, but not a perfect speller; well-read writers usually have problems with a tiny percentage of the words they write. I think that the solution to this problem is to let spelling develop naturally through massive reading in the early years, and provide older writers with some guidance in the use of spell-checkers and spelling dictionaries, as well as advising them to delay spelling concerns until the final draft. Good evidence that this is a reasonable solution is the fact that so many of us know when we are about to make a spelling mistake, and we can usually recognize the correct spelling of a word when presented with alternatives on a spell-checker. I think that this feel for correctness comes from massive reading. It is pointless to urge children to look up words before they have developed this "spelling sense." The article to which Stephen Krashen was responding: November 20, 2002 Studies Back Lessons In Writing, Spelling By Debra Viadero Education Week Adults may wince at painful childhood memories of penmanship lessons and spelling tests. A small but growing number of studies, though, suggest that systematically teaching handwriting and spelling might actually help some students write more and do it better. "Most kids who are developing as writers are planning their writing as their pen hits the paper," said Steve Graham, a special education professor at the University of Maryland College Park. "If you have to switch your attention to figuring out how to spell a word, for example, that disrupts your planning process." The new research, which comes out of work by Mr. Graham and others, follows recent decades that saw a de-emphasis on formal spelling and handwriting instruction in the classroom. A major thrust of the approaches of the past 30 years has been to save such instruction for "teachable moments" and embed it in real writing and reading activities. But the newer findings suggest that teachers might want to rethink those ideas-at least for the worst spellers and writers in their classes. The latest study in this body of research is a report slated to be published this year in the Journal of Educational Psychology by Mr. Graham and two colleagues, Karen R. Harris and Barbara Fink Chorzempa. As part of that experiment, the researchers culled 60 2nd graders with spelling problems from four schools in the Washington area. Half the children got extra lessons in mathematics, and half got the same amount of added instruction in spelling. At the end of 48 lessons, the children in the second group had improved more than their spelling. They also wrote more fluent, better-constructed sentences and texts than the children who had only gotten extra math lessons. A caveat here, however: The spelling-but not the writing-benefits held up when the children were tested again six months later. Building on the Best The study built on earlier work that Mr. Graham did with Virginia W. Berninger, an educational psychology professor at the University of Washington in Seattle. As part of that 1998 study, researchers also focused on the poorest spellers from 2nd grade classrooms. The 128 children were divided into seven treatment groups and taught each one different spelling- improvement strategies. While students in several groups became better spellers at the end of 24 lessons, one group also became more fluent writers. That group used an approach that combined several different strategies. Both Ms. Berninger and Mr. Graham, in similar kinds of studies involving handwriting, have documented the same phenomenon: Students with poor penmanship who are given handwriting lessons produce better, more fluent writing than counterparts who get no such instruction. Ms. Berninger emphasized that the kind of handwriting and spelling lessons she has been testing are far different from the traditional drills that adults remember from childhood. "Whenever we teach handwriting, we teach composing in the same lesson, and the same with spelling and reading," she said. "We've really tried to build on the best contributions of all of this research in brain research, process writing, and cognitive process." Both experts believe that early and explicit instruction in spelling and penmanship can stave off many problems later on in 3rd and 4th grade, when children tackle more complex writing tasks. What's more, they add, educators shouldn't expect computers to solve all of their students' "text transcription" problems. Keyboards can be slow going for beginning writers, they note. Studies also show that computerized spell-checkers fail to catch about half of students' misspellings. Learning to read Links to Education Resources HOME