Psychol Res (1981) 43:245--257 (Developmental Dyslexia Issue)
Psychological Research © Springer-Verlag- 1981
Spelling, Reading and Adult Illiteracy Dolores Perin MRC Developmental Psychology Unit, Drayton House, Gordon Street, London WC10AN, England Summary. This is a study of the relationship of spelling to reading in adults. The spelling of six adult literacy students who read well or poorly was analysed to discover whether error patterns resembled those previously reported for children. Three tasks were administered, induding dictation and free writing of real words, and dictation of nonsense words. Good readers made many more phonetic errors than poor re~.ders did, indicating that their cognitive processes in spelling are similar to children's. In the nonword task, poor readers were less able than good readers in translating phonemes to ~ranhemes. It is argued that implicit knowledge of the relationships of letters to sounds provides a strategy for dealing with unfamiliar written material and it is in this process that poor readers are impaired.
A study of standards of reading in eleven- and fifteen year-olds carried out by Start and Wells in 1972 suggested that many young adults leave school each year with a standard of reading that is not adequate for normal functioning in a literate society. In the last five years many local authorities around Britain have set up literacy schemes for adults who are dissatisfied with their standards of literacy (Jones and Charnley 1978; Hargreaves 1980). Among these adults, spelling ability is always a problem. In many, reading also needs improvement but there is a fair number of adult literacy students who read adequately or verv well but find spelling an almost impossible task. Put in another way, it can be said that among adult literacy students, there are two subgroups of poor spellers, those who read at least to a functional level and those who read poorly. Since a number of studies h,xs found an effect of reading ability on strategies used in spelling in children, it is interesting to investigate spelling in adult literacy students, who show dear-cut ability differences in the two skills. A comparison of spelling in adults and in children is important because it is possible that they both experience the same problems. Organizers and teachers on adult literacy courses have worked hard to create and preserve a learning atmosphere which is non-threatening to students (Mace 1979). The main concern is to help students overcome feelings of inadequacy caused by their lack of the 0340-0727/81/0043/0245/~ 02.60
246
Dolores Perin
basic skills that are usually taken for granted in a literate society. Assessment and testing which might lead to a repetition of the experience o f failure which many students had as school children is usually avoided in adult literacy classes. Research into cognitive processes underlying the reading and spelling o f students is not encouraged, as this often invloves assessment through testing. However, the author participated in an adult literacy scheme and permission was received by organizers, teachers, and students for research into spelling to be carried out (Perin 1980). In one set of studies, six men who were receiving individual tuition in literacy with volunteer tutors agreed to take part. Subjects were aged from the early twenties to the late fifties and were of low socio-economic status. While life history data which could provide an explanation for reading and spelling difficulty in these individuals was not collected, it should be noted that economic and cultural underprivilege, childhood absences from school due to illness or truancy, and family mobiltv often occur in conjunction with illiteracy (Hunter and Harrnan 1979) However, the lack of clear-cut causation in these subjects should not be a deterrent to understanding their stren~xhs and weaknesses inspelling and reading. Testing took place individually either in the student's or the experimenter's home, following the procedure in individual literacy tuiton that was followed at the time. Students were given complete feedback on their performance and were apprised of the reasons for testing. Although the tasks were quite difficult for several students, all found the experience worthwhile. In a preliminary assessment of reading, a 100-word-long passage (consisting mostly of high frequency words) from a booklet on job retaining opportunities for the unemployed, produced by the Government, was presented for oral reading (see Appendix I). There was a clear difference in subjects: three of the students read fluently with a minimum of errors and three read extremely slowly, hesitating over many words and making many errors. All six subjects found spelling difficult, which was a major reason for enrolling for literacy tuition. In the first experiment, spelling ability was assessed and an attempt was made to discover whether reading skill affected strategies used in spelling. The passage which subjects had read was used to investigate spelling. Since spelling was such a problem for subjects, to avoid undue anxiety, the test was first presented ,for copying. Although the poorer readers took longer, only one of the subjects produced more than one error in copying. This subject, a poor reader, made seven errors in copying. Six of the errors involved the omission of one letter and one error involved the substitution#f one letter. The single errors made by the other subjects were of the same types. The passage was then dictated. Each whole sentence was read aloud by the experimenter, then phrases, and finally individual words. Subjects were told that if they were not sure of the spelling of a word, they should write the best ap0roximation to it. Despite the fact that they had read and then copied the passage, all subjects made many errors on the dictation task. Table 1 shows reading scores and spelling (dictation) scores for each subject. Scores are expressed as percentage of words correct. Appendix II contains the corpus of errors made in dictation. As expected from preliminary assessment, there was a large difference in reading ability between the two groups of three subjects. Also, while spelling was impaired in all subjects, the problem was more serious in the poor readers. One way to view the groups is to describe the good readers as showing a reading-spelling discrepancy and the
247
Spelling, Reading and Adult Illiteracy Table 1. Experiment 1 : reading and spelling scores (percentage of words correct) Good readers Reading
Subject
Group mean
100
Subject
99.0 (1.53)
Reading
Group
87 79
Spelling Group mean
mean
66
97 99
Poor readers
Spelling
89 77.33 (10.60)
92 80
Group mean 51
87.0 (6.24)
65
60.67 (8.39)
66
poor readers as nondiscrepant. Spelling problems are 'unexpected' (Frith 1980) in the good readers while they are not surprising in the poor readers, given the relatively high correlations which have been found between reading and spelling (e.g., Horn 1969). The present findings indicate parallels between children and adults. Frith (1981) has described children who read well while spelling poorly as 'dysgraphic' while children both reading and spelling poorly may be termed 'dyslexic'. This categorization, which is helpful in distinguishing two specific types of problem occuring in the development of literacy, may also be a useful way of looking at adult subjects. One major difference between dysgraphics and dyslexics was spelling error pattern. While children who read better than they spell make a predominance of phonetic spelling errors, children whose reading and spelling are equally poor make many nonphonetic spelling errors (Frith 1978; Nelson and Warrington 1874; Sweeney and Rourke 1978). Since reading and spelling are differentially impaired in the present adult subjects it is possible that strategies used in spelling differ in the two groups. An analysis was made of the spelling errors produced in the dictation task. All errors were examined except functors and correctly spelled homophones. Where a word was erroneously spelled twice, only the first misspelling was analysed. Each error was categorized according to whether the pronounciati0 n of the target had been retained. The purpose of this analysis is to discover whether the two groups differed in the strategy they employed in spelling, when subjects were not sure of the word's orthography. Phonetic errors were words for which the overall sound frame of the target had b e e n preserved even if one of the component phonemes happened not to be correctly represented. Scoring was lenient in allowing for regional variation in pronounciation. Phonetic errors were expressed as a percentage of each subject's total errors. In this way, error quality was clearly distinguished from error frequency and the question could be asked whether poor readers made not only more spelling errors than good readers but whether they made different types of errors. Table 2 shows the percentage of phonetic error for each subject. A large difference between the two groups in error pattern was found. Table 3 shows examples of phonetic and nonphonetic errors made by subjects in dictation. The predominance of phonetic spelling errors among the better readers and of nonphonetic errors among the poorer readers suggests that the two groups are using different strategies in spelling unfamiliar words, i. e., words whose complete orthographic
Dolores Perin
248 Table 2. Experiment 1 : percentage of phonetic spelling error Poor readers
Good readers Subject 1
2 3
% 48 85 62
Group mean
Subject
65.0 (18.68)
%
Group mean
8 24 25
19.0 (9.54)
Table 3. Experiment 1 : example of phonetic and nonphonetic spelling errors
Target
unemployed particular industry different project
Response Phonetic
Nonphonetic
uninployd pertickular industrey, indistry diffrent, differant progect
unyep, unpalled, annepld, unely partqulr, petikle inmtree, industy, indstre difort, differt, deifrut poragced, pojet, prget
structure is partially or wholly unknown to subjects. Just as in the studies with children it appears that good readers are using a phonological route, i.e., phoneme-grapheme translation rules, in spelling, while this route is either impaired in poor readers or t h e y are attempting to use a lexically based strategy which may or may not be visual in nature. However, the difference in error pattern m a y be specific to a dictation task and in normal writing a different pattern of errors may be found. Nelson (1978) suggests that comparisons o f spelling errors in good and in poor spellers may be confounded b y the fact that, for poorer spellers, the words m a y be t o o difficult, and that this would affect the quality of error. It is possible that, when faced with the task of spelling a word which the subject has hardly ever seen or spelled before, subjects may just make a wild guess. Poorer readers, having smaller reading vocabularies, would make more wild guesses than better readers would. However, if subjects generate their own meanings, it is more probable that the words they would choose would be within the range of their reading vocabularies and would not be completely unfamiliar, and could therefore be attempted b y means o f a phonemic strategy. Experiment 2 was conducted to study error patterns in a free-writing exercise. It is assumed that in the course o f free writing, when subjects self-generate meanings, words will be chosen which are either within or just beyond their ability range. It seems unlikely that subjects would use words which they find extremely difficult and at which they might have to guess wildly. Accordingly, a task was designed in which subjects were asked to write brief stories on the basis of a story frame. The story used was based on a plot found in a popular comic. Although in one sense this is an easier task for subjects who can choose the words which t h e y spell, in an important sense it is a stricter test of the supposed cognitive strategies used in spelling because the text is not read or copied beforehand.
Spelling, Reading and Adult Illiteracy
249
Table 4. Experiment 2: performance in free writting
Story : No. words Subject
Group
Good readers % % Story: correct i phonetic error No. words Group Group Subject
] mean
1
2 3
46 78.33 1108 (31.09) 81
mean
24 91 88
mean
57 67.67 (37.85)
63 75
4 65.00 (9.17)
Poor readers % % correct phonetic error Group
Group
Group
mean
mean
mean
68
5
75
6
66
40 69.67 (4.73)
6 62.00 87 (23.64 5O 59 33
Subjects were seen individually. The experimenter read out a story frame twice with repetitions as requested (see Appendix III) and subjects were told to write the story in their own words, changing or elaborating it as much as they wished. Fifteen minutes were allowed to write the story, and more time was given if requested. Subjects were told not to be unduly concerned about making spelling errors; if they were unsure of a spelling, they should guess as well as they could. Appendix IV is the corpus of errors produced by each subject. Table 4 shows story length, percentage of words spelled correctly, and percentage of errors which were phonetic, for each subject. There is an interesting difference in performance between the dictation and freewriting tasks. While spelling ability differed considerably between good and poor readers in dictation (group means were 77.33% and 60.67% respectively), spellers' ability as assessed by free writing was not as different between the two groups (67.67% and 62.00% respectively). However, this may be accounted for by subject l's poor performance in free writing compared with dictation. His reading and spelling.abilities are the most discrepant and he might have made more dictation errors had he not seen and copied the passage beforehand. It is considerable that this subject benefited especially from the initial copying task. The two groups continued to differ in their percentages of phonetic spelling error (65 % as compared with 29.67%). An interesting comparison may be made between subject 3, a good reader, and subject 5, a poor reader. Spelling ability as assessed by free writing was almost identical but the good reader made 75% phonetic spelling errors while the poor reader made 50 %. The trend for better readers to make phonetic spelling errors and for poorer readers to make nonphonetic errors reflects a difference in cognitive processes underlying spelling. It is important to make the distinction between strategies used when a spelling is familiar and those when it is unfamiliar. In spelling words that they know, good and poor readers may not differ very much in the strategies they use. In contrast, a qualitative difference ist found between good and poor readers where unfamiliar words are concerned. Here, strategies would differ between the groups, as reflected in their spelling error patterns. Simon and Simon (1973) discussed the kinds of errors that might result from attempts to spell unfamiliar words. According to their model, spellers will in this case attempt to spell syllable by syllable. It is probable that a phonemic strategy is usedhere. If spelling of a syllable is unsuccessful, an attempt is made to analyse its phonemic struc-
29.67 (22.19)
250
Dolores Perin
ture and then apply phoneme-grapheme translation rules. It should be emphasized that the speller may or may not be conscious that he is 'applying rules': this is normally an implicit or unconscious process. The speller generates phonemic approximations to the word until the spelling 'looks right.' The spelling error patterns found in the dictation and free writing tasks suggest that good readers use such a phonological route in spelling unfamiliar words whereas poor readers make less use of this strategy. Experiment 3 was carried out to discover whether even poor readers could be induced to use a phonemic strategy in spelling or whether they in fact have a basic inability to use this route, as reflected in their preponderance o f nonphonetic real word errors. This experiment was a dictation task. Twenty-eight nonsense words were constructed on the basis o f initial and terminal consonant clusters appearing in the 800 most frequent words in written English (KuEera and Francis 1967). The list of stimuli in the order of presentation is shown in Appendix V. All words were monosyllabic with the exception of four. Subjects were asked to spell some 'made-up' words. It was emphasized that there was no 'definite right or wrong spelling' but that subjects should try to spell them so that another person reading them could pronounce them in the same way as they had been dictated. This emphasized the need for a phonological strategy. The subjects were told to watch the experimenter as she pronounced each word. The word was said twice and then subjects wrote their responses. Further repetitions were made if requested. Subjects were seen individually. Deliberately, no analysis of the spelling of vowels was undertaken, as it is known that vowels present a particularly difficult problem in English. Seymour and Porpodas (1980) observed a preponderance of vowel errors in the spelling of nonsense words. Only the first two and last two consonant graphemes in each nonsense word were scored, that is, as individual phonemes. A response was scored as correct only if the sound of the target consonant phoneme was accurately represented, irrespective of how the rest o f the word was spelled. The corpus o f transcriptions is shown in Appendix VI. Table 5 shows the number of errors made for each subject, expressed as a percentage o f the phonemes scored. The poor readers made many more errors than the good readers did, indicating that the better readers had a superior knowledge and use of rules for spelling. There was no overlap between the two groups. However, if the results of nonsense word transcription are considered in relation to percentage of phonetic real word errors, it seems that all subjects showed that they were capable of using a phonological strategy in spelling nonwords. While a direct numericalcomparison is not
Table 5. Errors made in Experiment 3 nonsense words) Good readers Subject
Poor readers
% error
Subject
% error
Group mean 1 2 3
16.30 6.12 13.63
12.01 (5.28)
Group mean 4 5 6
58.61 29.56 24.17
37.45 (18.53)
Spelling, Reading and Adult Illiteracy
251
possible, owing to different scoring methods, this finding is particularly interesting because it indicates a greater knowledge of the rules than was actually shown in realword spelling. The transcriptions of the nonsense words indicate specific differences among the subjects. Two of the poor readers (subjects 4 and 5) made many errors involving liquid and nasal consonants in both initial and terminal clusters. This may be due to a lack of conscious representation of these particular phonemes in certain phonetic contexts (Marcel 1980a). For attempting to help poor spellers to improve their skill, tasks such as the present one can give important information of particular component processes that require improvement. As Marcel points out, the phonics approach (McCaffery 1979, provides an interesting discussion of the use of phonics in adult literacy tuition) may not be an adequate learning method for students who do not internally represent particular phonemes in the normal way. In accounting for the difference in performance found in the three tasks between good and poor readers, it is possible that the use of phoneme-grapheme rules is linked specifically to the development of spelling ability rather than reading. Bryant and Bradley (1980) have argued that spelling (but not reading) is originally carried out through a phonological strategy and that it might be implicit learning of the rules at this early level that accounts for the difference. This appears plausible because in general, reading and spelling levels tend to be strongly associated: good spellers, who learned the rules during acquisition phases are also good readers. However, Frith (1979) found that subjects matched on spelling whose reading differed showed different spelling error patterns, indicating an effect of reading on spelling. This seems possible in the present sample because in Experiment 2 although spelling error frequency did not differ very much, error pattern differed considerably between good and poor readers. Advancement in reading is possible when single words can be read accurately (Shankweiler and Liberman 1972). Children who read well are superior to those who read poorly when reading nonsense words (Frith 1972, Perfetti and Hogaboam 1975; Snowling, this issue), an ability which requires grapheme-phoneme translation, This suggests that advancement in reading depends on the ability to use the specific rules of written English. (Knowledge of these rules may of course be implicit.) Although rules used for reading and those used for spelling are not entirely reversible (Henderson and Chard 1980), the notion that letters and sounds are related in a lawful way provides a useful strategy for both production and reception of written language. T,he incidence of phonetic errors in the spelling of good readers reflects their ability to use phoneme-grapheme rules when they need to spell an unfamiliar word. On the other hand, since poor readers make many nonphonetic errors in spelling, it appears that these subjects have not achieved an adequate grasp of the rules, However, all subjects, both good and poor readers, showed a greater use of phoneme-grapheme translation in nonsense word than in real-word spelling. This is no doubt because in spelling real words, reference may be made to stored lexical images which may be visually based in memory. Therefore spelling unfamiliar real words could involve a combination of auditory and visual strategies. On the other hand, lexical access is not possible for nonsense words themselves (although lexical access for words by analogy is possible along the line suggested by Glushko 1979b, and Marcel 1980b) and reliance on phoneme-grapheme rules is more necessary. Although all subjects used phoneme-
252
Dolores Perin
grapheme rules in the nonsense word task more than might have been expected from their real-word phonetic errors, the differences were considerably larger for the poorer readers. These subjects appear to be using different strategies depending on whether stimuli are real or nonsense words. By inference it is probable that poorer readers tend to have a 'visual' approach to written language (Rozin and Gleitman 1977). They-do not exploit the fact that alphabets provide guidelines to pronunciation and therefore are disadvantaged in approaching new written words. Nonphonetic spelling errors may result from random guesses, inefficient use of phoneme-grapheme translation rules, or inappropriate 'lexical images.' The spelling o f the poorer readers in the dictation and free-writing tasks may reflect their preference, when there is a choice, o f nonphonetic (and possibly visually based) processes. However, when there is relatively little chance to use a lexical strategy, as with nonsense words, poor readers demonstrate that t h e y can in fact use the rules more accurately than expected. This has implications for the remediation of spelling in poor readers: they could be encouraged to use their knowledge o f p h o n e m e g r a p h e m e rules in generating spellings. While the characteristics of English orthography often prevent phonetic spellings from being the correct ones, spellings generated on the bails of the rules can at least be understood b y recipients o f written messages. Furthermore, use o f the phonological route may well be a necessary stage in the development o f spelling.
A p p e n d i x h Text Used to Assess Reading and Spelling If you are aged 16-19 and unemployed you should take advantage of the special training schemes run by the government for unemployed young people. Enquire at your local Jobcentre about the different schemes available. You can choose to work for an employer on the spot to get experience of a particular type of job or you can work on a special project. Or you may prefer to work in Community Industry. There are also courses run to help you choose which kind of work suits you best and courses to train you for a particular job at operator or semi-skilled level. (From: Fresh Start: A Guide to Training Opportunities, p 9. Published by the Equal Opportunities Commission, April 1978)
Appendix II: Errors Produced in Dictation Response
Target Subject 1 aged unemployed should take advantage special training schemes run
by
agad unpalled, uninepould shoud
Subject 2
Subject 3
shoud
avdvantch spaech, space
addvantage
soirees
sceance
specele, specel scemes
Subject 4
Subject 5
ace unyep unepyld shud
age unely, uneyed
efceh spell
adventagus spesll, speasll traing scemes
traning sarms, scem wand bary
Subject 6 annepld, anninpeld choud tack evtge sechcul trying skmen, scimze ren
SpeUingl R e a d i n g and A d u l t Illiteracy
253
A p p e n d i x I h Errors P r o d u c e d in D i c t a t i o n (cont.) Target
government young people enquire your local jobcentre different available choose an employer spot experience particular type or job project prefer community industry also courses
Response Subject 1
Subject 2
govment
govermant
peple equire
enquier
Subject 3
pepole enzuiare
Subject 4 commat lan pepo inqoor
Subject 5
Subject 6
govement youg pleople inequr
guvmut yae peal enqe
locd jobcenter differt avilable
loklon jobcenter deifrut avelbol tues and enplore sopte exsprens petylas, petikle tipe are
yo ure
locke jobscenter diffrent avabloue chouse
johsenter differant avalibale
epoury exectants pertickular, perticule tip our
avalible chuse
experiance
profore comunty industrey coursecs
lokel jodcanter difort aboydall cosi, case and emplene expuntes ptorer
ely shot exsprat partqulr
tiep
tip
progect profer comunite indistry
jod pgett prae commert inmtree
projet perfer commet industy
corses
callste,
corses
enployer
poragced
cousses
johcentor
experiance
prget brfere cumnete indstre olso crse, corse
coats
which kind suits best train operator semi-skilled level
Appendix IIh
ciund suets
hoperate semhem livle
Story F r a m e
leval
wick caman suite stuten bust traine trime opater semey-skilled semskill lad
. kid suat
cider sune
tegne operater opratte same s smesced semeslde leavl levle
f o r F r e e Writing Task
There was a very big diamond, one of the biggest diamonds in the world. It was being shown in a museum. The museum was always crowded. One day, there was a bomb scare. Everyone had to leave the museum, even the men guarding the diamond. When the guards returned, the diamond was gone. It turned out that thieves pretending to be bomb experts had stolen it.
254
D o l o r e s Perin
A p p e n d i x IV: Errors P r o d u c e d in F r e e Writing
Subject 1 Target
Response
Target
diamond museum were lots there bomb scare disposal they
dirmend mouceme where loess their born skey disposaile there, thaire takeon
scare guards taken returned back gone thief stolen by disposal search
taken
Subject 2 Target Response diamonds them their surprising prepared stealing fraud demand something iron
Target
Subject 3 Response
museum biggest while people somebody, bomb their investigate they thieves
Target
dimonds the there serpising prepeared stelling froud demarnd somethink lone
muusem bigest wile pepole somerbody bome there investegate thy thifeves
Subject 4 Response
there diamond museum looking lots people bomb
the droman nemcone look lasts peaper bonod
Target
Subject 4 (cont.) Response sehoec garase tanr ratand blant gornd thefe stem, scoland bay basor sertc
Subject 5 Response
diamond museum people busy that holiday biggest bomb scare
dimon muiu, mimus pleople bise anthe holday biges bown schere
Subject 6 Target
Response
diamond museum large lots people bomb
dimennd, dimemd muszerm lege lost peple borne
scare
serae
everyone told leave building they guards check saw been stolen by thieves who pretended experts
everone teld live bleing that gres cheke sore he solnd be thesve how brtende exspes
Spelling, Reading and Adult Illiteracy
255
A p p e n d i x V: N o n s e n s e W o r d D i c t a t e d in E x p e r i m e n t 3 Phonetic transcription
Possible spelling
1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
plard fract, fracked threng shurth stigle, stiggle plunce, plunts smenk schoond, scooned prolp granst, granced thurl, thirl breld strable, strabble trult blict, blicked stoft, stoffed sparn squant freple, frepple shotle, shottle preft, preffed plost, plossed stuch freth blurk, blirk prunge sturld, stirled chirst, cherst
p 1~:rd ~"r "~ I
~ra:nsf
03:rl 12. bre Id 13. s t ~ b l 14. t r a i l 15. hiz gt 16. sT~:~t 17. ~ p ~ : r . 18. s ~w 0r~1: 19. {~re~l 20. S ~ t I 21. pre;'~ 22. p I~ s t 23. ~ t A t S 24. ~r e 6 25. bt~:rk 26. prAnd5 27. ~ ~ 3: r Id 28. t ~ 2 : r g t 11.
A p p e n d i x V I : R e s p o n s e s in N o n s e n s e W o r d D i c t a t i o n T a s k Target Subject plard fract threng shurth stigle plunce smenk schoond prolp granst thurl breld strable trult blict
Response Good readers 1 2
3
4
Poor readers 5 6
plard frate frang shouth sraull pluns smenk skund proup gruset thural brund strubell trultt bleact
plard fract pregn eherf siggel plunce cermenck eoond prolp grarnsd thurl breld starble trait blit
pralt frake thandeg shut stodo plate send seholl ploged gent fr belled seding tomt bldord
plard frak thag shrth stigal pluts spek scood poop grst thural blld stabl trout bilt
plared fracked thrage sherth stigal plunse smanck skoond prolbe grunst theril dreld strable trount blicked
pelod frate thraing shere sigel plunge sgound propt grnst thele breld strabld trote blites
256
Dolores Perin
Appendix VI: Responses in Nonsense Word Dictation Task (cont.) Target Subject stoft sparn squant freple shotle preft plost stuch freth blurk prunge sturld christ
Response Good readers -1 2
3
4
Poor readers 5 6
stuft spound squnt fropall shoutiU preft ploust stusch throth biuak pracnch stally chosch
stoff spare cazant fropal schatal preft plarst stuch preff blurk prung surled cherced
stert spande sronte fro stealey pleats plones stening tiered dronck peace steared cheend
stoff spren squot thaple shotel paff plost stuch thaff blrk puj stred jrch
stoffed sparn squont frepale shotale prefet plost stuck threth blerck prunge sterled churst
stoftse sprne sjonte frabol shotol preft ptost stathes frthe plkets prnge sdd cherte
Acknowledgements: This research was undertaken while I was a D Phil student at the University of Sussex, funded by an SSRC studentship. I am grateful to Dr B. Lloyd for her encouragement, and to Dr U. Frith for her comments on earlier drafts of this paper. I would also like to thank the students, tutors, and organizers in the Brighton Literacy scheme who helped make this study possible.
References Bryant PE, Bradley L (1980) Why children sometime write words which they do not read. In: Frith U (ed) Cognitive processes in spelling. Academic Press, London Firth I (1972) Components of reading disability. Unpublished doctoral dissertation, University of New South Wales, Kensington, NSW, Australia Frith U (1978) From print to meaning and from print to sound or how to readwithout knowing how to spell. Visible Language 12:43--54 Frith U (1979) Reading by eye and writing by ear. In: Kolers PA, Wrolstad ME, Bouma H (eds) Processing of visible language, vol I. Plenum, New York Frith U (1980) Unexpected spelling problems. In: Frith U (ed) Cognitive processes in spelling. Academic Press, London Frith U (1981) The similarities and differences between reading and spelling problems. In: Rutter M (ed) Behavioral syndromes of brain dysfunction in childhood (in press) Glushko RJ (1979b) The organization and activation of orthographic knowledge in reading aloud. J Exp Psych Hum Percept Perform 5:674--691 Hargreaves D (1980) Adult literacy and broadcasting: The BBC's experience. Frances Pinter, London Henderson L, Chard J (1980) The reader's implicit knowledge of orthographic structure. In: Frith U (ed) Cognitive processes in spelling. Academic Press, London Horn TD (1969) Spelling. In: Ebels RL (ed) Encyclopedia of educational research, 4th ed, Macmillian, New York
Spelling, Reading and Adult Illiteracy
257
Hunter CStJ, Harman D (1979) Adult illiteracy in the United States. McGraw-Hill, New York Jones HA, Charnley AH (1978) Literacy discussion: literacy and adult education: the U.K. experience. International Institute of Adult Literacy Methods, Teheran Ku~era H, Francis WN (1967) Compul;ational analysis of present-day American English, Brown University Press, Providence, RI McCaffery J (1979) Using phonics. Adult Literacy Unit Newsletter No 7, Oct-Nov 1979
Mace J (1979) Working with words: literacy beyond schooling. Chameleon, London Marcel AJ (1980a) Phonological awareness and phonological representation: investigation of a specific spelling problem. In: Frith U (ed) Cognitive processes in spelling. Academic Press, London Marcel AJ (1980b) Surface dyslexia and beginning reading: a revised hypothesis of the pronunciation of print and its impairments. In: Coltheart M, Patterson KE, Marshall JC (eds) Deep dyslexia. Routledge & Kegan Paul, London NelsOn HE (1978) Spelling development and memory functions in childhood dyslexia. Unpublished PhD thesis, Faculty of Science, University of London Nelson HE, Warrington EK (1974) Development spelling retardation and its relation to other cognitive abilities. Br J Psychol 65:265-274 Perfetti CA, Hogaboam T (1975) The relationship between single word decoding and reading comprehension. J Educ Psychol 67:461-469 Perin D (1980) Spelling difficulty in school leavers and adults. Unpublished DPhil thesis, University of Sussex Rozin P, Gleitman LR (1977) The structure and acquisition of reading II: The reading process and the acquisition of the alphabetic principle. In: Reber AS, Scarborough DL (eds) Toward a psychology of reading. Erlbaum, Hillsdale, NJ Seymour PHK, Porpodas CD (1980) Lexical and non-lexical processing of spelling in dyslexia. In: Frith U (ed) Cognitive processes in spelling. Academic Press, New York Shankweiler D, Liberman IY (1972) Misreading: a search for cause~s.In: Kavanagh JF, Mattingly IG (eds) Language by ear and by eye. MIT Press, Cambridge, Mass Simon DP, Simon HA (1973) Alternative uses of phonemic information in spelling. Rev Educ Res 43:115-137 Snowling MJ (1981) Phonemic deficits in developmental dyslexia. Psychol Res 43 (this issue) Start KB, Wells BK (1972) The trend of reading standards. NFER, Slough, UK Sweeney JE, Rourke BP (1978) Neuropsychological significance of phonetically accurate and phonetically inaccurate spelling errors in younger and older retarded spellers. Brain Lang 6:212--225