Why spelling reform? A review of the reasons with a focus on Danish

Min bacheloropgave.

English resumé:

In this thesis, I discuss language reform with a focus on orthographic/spelling reform. I briefly review the some prior cases of language reform. I then discuss the motivations for and against reform spellings. For this I draw upon work from psycholinguistics, educational linguistics (phonics), experiments with simplified spelling systems and cross-country comparisons. Then, I review the Danish orthography and its problems. Finally, I review my proposal for a Danish spelling reform, Lyddansk.

Dansk resumé:

I denne opgave diskuterer jeg sprogreform med et fokus på ortografisk/stavemæssig reform. Jeg gennemgår kort nogle tidligere tilfælde af sprogreform. Derefter diskuterer jeg grundene for og imod en reform af stavemåderne. I min diskussion inddrager jeg værker fra psykolingvistik, uddannelseslingvistik (phonics), eksperimenter med simplificerede stavesystemer og tværnationale sammenligninger. Efter det så gennemgår jeg dansk retskrivning og problemer med den. Til sidst, så gennemgår jeg mit forslag til at reformere danske stavemåder, Lyddansk.

Opgaven er på engelsk.

bachelor lingvistik spelling reform

Reversal-testen og morfemkonstans

I min bachelor diskuterer jeg brugen af reversal-testen til at spotte status quo bias. Jeg skriver:

3.3.5 Status quo bias

A number of cognitive biases have been discovered by now (e.g. Kahneman, 2011). A well-studied one is status quo bias, which is the tendency for people to favor the current system over alternatives, regardless of what the current system is. They do this by rationalizing their preference for the current system over the alternatives.

The question is how we can rationally figure out when someone, perhaps ourselves!, is affected by status quo bias. The philosophers Nick Bostrom and Toby Ord (2006) proposed the reversal test to solve this. They write:

Reversal Test: When a proposal to change a certain parameter is thought to have bad overall consequences, consider a change to the same parameter in the opposite direction. If this is also thought to have bad overall consequences, then the onus is on those who reach these conclusions to explain why our position cannot be improved through changes to this parameter. If they are unable to do so, then we have reason to suspect that they suffer from status quo bias. (p. 664-665)

The evidence from the reversal test is defeasible, i.e. if they can supply reasons to think that we are in a local optimum, it can be reasonable to disfavor any change in a parameter.

If we return now to the etymological argument above, we can see that it is plausibly a case of status quo bias. To do this, simply ask proponents of that argument whether they would prefer us to return back to even more etymological spellings. In Danish, we now write <suveræne> (sovereign) instead of the more etymological <souveraine> (as found in the King Law of 1665). Yet etymology was offered as a reason to prefer the current <mayonnaise> spelling over the Danified <majonæse>. Alternatively, consider the word <løjtnant> (lieutenant) which used to be spelled <lieutenant>. Should we go back?

Jeg nævner etymologi (nogle gange ligefrem kaldt et princip). Men en samtale i dag med kæresten gav mig et eksempel på hvordan morfemkonstans ikke er brugt konsistent, og nok heller ikke ville blive.

Det drejer sig om ordet <handske>. Ordet er tydeligvis beslægtet med ordet <hånd>, men både udtalen og stavemåden er forskellige. Stavemåden er her rimelig fonemisk (man kunne diskutere om <hond> ville være bedre, men det er fint givet det nuværende danske stavesystem). If. morfemkonstansprincippet, så burde morfemet <hand/hånd> staves ens på trods af at lyden er forskellig (som i fx <god> og <godt>). Mit spørgsmål til personer som er fan af dette princip er således om de støtter en ændring af et af ordene så der blir morfemkonstans? Hvis nej, er der så en god rationel forklaring på hvorfor ikke? Hvis ikke, så har vi god grund til at tro at osse morfemkonstansprincippet bruges som efterrationalisering i stedet for en rigtig rationalisering.

Studie: Indsaml fejlrettelsesdata fra Wikipedia

I forbindelse med noget andet jeg var ved at arbejde på, så stødte jeg på et paper: Pfeil, Ulrike, Panayiotis Zaphiris, and Chee Siang Ang. Cultural differences in collaborative authoring of Wikipedia. Journal of Computer‐Mediated Communication 12.1 (2006): 88-113. Selve hoveddelen af artiklen er ikke voldsomt interessant eller særlig overbevisende. Men det deres data gav mig en god ide:

wiki_comparative

Jeg kikkede naturligvis mest på de sproglige, grammar, spelling. Jeg tænker på en tabel fra Foundation literacy acquisition in European orthographies, British Journal of Psychology (2003), 94, 143–174.

foundation lit

Af de fire sprog de har valgt, så er fransk det sværeste if. tabellen, hvorefter at man skulle forvente flere stavefejlrettelser. Derefter hollandsk, og til sidst tysk. Det er præcis den rækkefølge med ser i dataene fra deres lille Wikipedia-studie. Japansk har også en lav stavefejlrettelses%, så spørgsmålet er naturligvis nu: passer det med hvor regulær den japanske retskrivning er? Jeg kikkede lidt omkring på Wikipedia, men kunne ikke umiddelbart finde noget om hvor regulær japansk retskrivning er. Men det er velkendt at de har et besværlig skriftsprog med 3 forskellige systemer af skriftegn.

Derefter gik turen til Google Scholar, hvor man kan finde alt muligt godt.

Hino, Yasushi, et al. “The effects of polysemy for Japanese katakana words.” Reading and Writing 10.3 (1998): 395-424.
Because each katakana character corresponds to a single syllable (mora), katakana is considered to be a shallow orthography which has virtually no spelling-to-sound irregularities. Thus, in terms of the dual-route model, the nonlexical route would be able to produce correct phonological codes for all these words. As noted, according to Balota and Chumbley’s arguments, if both word frequency and polysemy effects are due to lexical selection, both of these effects should not vary in size across tasks in which lexical selec- tion is fully involved. Our use of a completely regular orthography, however, changes those predictions. First of all, with respect to word frequency effects, the cross-task equivalence would not be expected because, as noted, a dual- route analysis suggests that low frequency words often do not require lexical selection. Thus, the expectation is that there would be a smaller frequency effect in naming than in lexical decision. More importantly, with respect to polysemy effects, the cross-task equivalence should hold for high frequency words because, for these words, the lexical route generates phonological codes much faster than the nonlexical route, meaning that the contribution of the nonlexical route to performance in the naming task would be minimal. For low frequency katakana words, however, the expectation would be that the polysemy effect should be smaller in naming than in lexical decision because of the large contribution of the nonlexical route in naming. [min emfase]
Delattre, Marie, Patrick Bonin, and Christopher Barry. “Written spelling to dictation: Sound-to-spelling regularity affects both writing latencies and durations.” Journal of Experimental Psychology: Learning, Memory, and Cognition 32.6 (2006): 1330.
The dual-route model of spelling production (e.g., Ellis, 1982) proposes that two processing systems operate in parallel: a lexical route that retrieves spellings of known words from a memory store of word-specific knowledge and a nonlexical (or assembled) route that generates spellings using a process of sublexical sound-to- spelling conversion. The assembled spelling route would be effi- cient in languages whose orthographies have predictable or con- sistent orthographic-to-phonological correspondences (such as Turkish, Italian, and Japanese kana) but would be considerably less effective for English and French, whose orthographies are characterized by highly inconsistent relationships (e.g., the vowel /i:/ is spelled in many different ways in English words, as in eel, tea, theme, thief, Keith, people, me, key, quay, ski, etc.). There are many irregular and some almost arbitrarily spelled words in En- glish (e.g., pint, yacht) and French (e.g., fraise, monsieur). The lexical route would work for all known words (irrespective of regularity) but could not provide spellings for new words or nonwords. The assembled route would work for nonwords but would often produce phonologically plausible errors (PPEs), par- ticularly to irregular words, such as yacht (YOT) and monsieur (MESSIEU). [min emfase]
Buchanan, Lori, and Derek Besner. “Reading aloud: Evidence for the use of a whole word nonsemantic pathway.Canadian Journal of Experimental Psychology/Revue canadienne de psychologie expérimentale 47.2 (1993): 133.

Reading Japanese Written Japanese consists of three distinct scripts. The logographic Kanji represents content words while the syllabic Kana scripts consist of Hiragana which represents grammatical morphemes and Katakana which represents borrowed words such as television and computer. Both Hiragana and Katakana have very consistent spelling-sound correspondences (i.e., they are shallow scripts). Transcribing a word which normally appears in one Kana script into the other Kana script produces a pseudohomophone, a word that is ortho- graphically unfamiliar at the whole word level but retains its original pronunciation. Since readers must rely entirely on the assembled routine to read such character strings aloud, any evidence of priming for these words is evidence that the use of assembled routine can result in priming. [min emfase]

Altså, uden at kende noget mere til japansk kana (katakana+hiragana), så lader det til at være meget regulært. Så skulle man forvente relativt få rettelser for stavefejl. Det er en skam at japan også bruger kanji, ellers ville deres retskrivning være næsten helt konsistent, hvilket ville få en til at forudsige, at de rettede færre fejl end tyskerne gør. Men deres fejlrettelses%er var ca. ens.

Næste skridt må jo være at kikke på nogle flere Wikipedia-sider for at bekræfte mønstret, og at kikke på andre sprog (især dansk og engelsk). Deres metode kræver dog en del manuelt arbejde, så den er ikke optimal. Måske man kunne forsøge sig med en keyword baseret løsning? Når man laver en rettelse på Wikipedia, så kan man nemlig skrive nogle få ord om hvad ens rettelse gik ud på. Mange som retter stavefejl skriver det nok her, hvilket gør det muligt at måle antal stavefejlrettelser udfra keywords. Det kræver dog stadig at man kan de sprog som man vil undersøge, ellers ved man ikke hvilke keywords man skal søge efter.

At gøre det ovenstående ville dog nødvendiggøre at dataene ikke var sammenlignelige med deres. Deres dataindsamlingsmetode er dårlig (kræver for meget tid), men ideen er god: bruge data fra Wikipedia til at estimere hvor ofte folk retter fejl.

Måske er variation ikke så slemt som man tror

Far be it from us to deny the handicaps stemming from the absence of a norm, but perhaps norms are in fact less important than we Westerners, conditioned by our very standardized written languages, might assume. After all, our own standardization is only a product of the eighteenth and nineteenth centuries; however effectively Alcuin of York may have prescribed for Latin, for the vernaculars each scribe and each publishing house had their own conventions in Europe until printing had been established for at least two centuries.

Living as they do in polyglot societies, pidgin speakers are accustomed to linguistic variation: rarely are they biased towards one particular spoken, still less one particular written, norm of pidgin. Possessing usually a rudimentary knowledge of the conventions of written English, they are tolerant of. or indifferent to, variant orthographies for pidgin: unlike the European reader they are not always unconsciously looking for the English etymon or related graphic form.

During the last parliamentary elections in Vanuatu, in December 1991. I was struck by the ease with which the voters of the Islands glanced over and understood the different election posters, some written by Anglophones, others by Francophones. Unfettered by any rule, each author had freely used his imagination to render the modem or technical terms that the voters had never seen written but had heard on the radio. All the scripts were equally accessible: might it not be our European norm that becomes a handicap? By constantly prohibiting any deviation in our languages, we are put off when any disparity appears in languages which should be very familiar.

Fra Vernacular Literacy: A Re-Evaluation (Amazon, Google Books) s. 233.

Ortografi, stavning, og læsning

En mere seriøs samling evidens for at det er dyrt at have en dårlig ortografi. Nedenunder findes en række videnskabelige artikler som viser diverse relevante ting.

—-

The impact of orthographic consistency on dyslexia: a German-English comparison
Cognition. 1997 Jun;63(3):315-34.
Landerl K, Wimmer H, Frith U.

Abstract:

We examined reading and phonological processing abilities in English and German dyslexic children, each compared with two control groups matched for reading level (8 years) and age (10-12 years). We hypothesised that the same underlying phonological processing deficit would exist in both language groups, but that there would be differences in the severity of written language impairments, due to differences in orthographic consistency. We also hypothesized that systematic differences due to orthographic consistency should be found equally for normal and dyslexic readers. All cross-language comparisons were based on a set of stimuli matched for meaning, pronunciation and spelling. The results supported both hypotheses: On a task challenging phonological processing skills (spoonerisms) both English and German dyslexics were significantly impaired compared to their age and reading age controls. However, there were extremely large differences in reading performance when English and German dyslexic children were compared. The evidence for systematic differences in reading performance due to differences in orthographic consistency was similar for normal and for dyslexic children, with English showing marked adverse effect on acquisition of reading skills.

Undersøgelsen drejede sig primært om ordblinde (dysleksiske) børn, men betragt også:
The main finding of the present cross-orthography comparison of developmental dyslexia was that the English dyslexic children suffered from much more severe impairments in reading than the German dyslexic children. The extent of the impairment varied according to the stimuli. For words of high frequency there was relatively little difference. However, for words of low frequency, the percentage of correct readings by English dyslexic children dropped to about 50% and remained as low for the long, three-syllable words, which tended to be also of low frequency. Nonwords posed even more severe problems for the English dyslexic children, despite lenient scoring. Here, the error rate for the three-syllable nonwords climbed to about 70%. In contrast, the German dyslexic children showed very few reading errors. Even for the long, three-syllable words their error rate was only about 10%, and for the three-syllable nonwords their error rate was only about 20%.
The enormous difficulties of the English dyslexic children with word and nonword reading were also reflected in very slow reading speed, for items which were correctly identified. The exceptions were highly familiar words which they presumably were able to recognize without having to rely on piecemeal phonological recoding. Thus, the differences between the English and the German dyslexic children were comparatively small for the short, high-frequency words (1.1 s vs. 0.7 s). For all other conditions, however, the differences were enormous. Even recoding of short one-syllable nonwords was twice as slow for the English dyslexic children (3.5 s vs. 1.4 s). For the few correctly read three-syllable nonwords the English dyslexic children needed as much as 6 s per item, compared to about 3 s for the German dyslexic children. This is not to say that German dyslexic children showed little if any deficits. Both the German and the English dyslexic children were significantly slower than age level control children not only
for nonwords, but also for words that could be considered as over-learned, such as short high-frequency words.
—-
Applied Psycholinguistics 24 (2003), 621–635
MIKKO ARO, HEINZ WIMMER
Fra diskussionsafsnittet:
The present study applied the pseudoword/number word/numeral reading procedure originally introduced by Wimmer and Goswami (1994) for comparing reading development of English and German children to several additional orthographies. The present results extend the original findings, the English–German replication of Landerl (2000), and the Spanish–Portuguese replication of Defior et al. (2002). They also provide additional important information on reading development in different orthographic contexts.
A key question was whether the minimal difficulty in pseudoword reading found for the young German and Spanish readers in the foregoing studies extends to other orthographies. This question is provided with a definite affirmative answer. Dutch, Swedish, French, and Finnish readers at the end of Grade 1 read the presented pseudowords with the same high accuracy as the German and Spanish children (between 80 and 90% correct), and some of these samples tended to improve further in the following grade levels. To evaluate these results, it has to be remembered that the pseudowords were presented in list format and the instructions stressed speed as well as accuracy. Therefore, some of the few errors may have resulted from reading too hastily. The conclusion is that the translation of new letter strings into acceptable pronunciations is easily acquired in all alphabetic orthographies involved in this study, with the exception of English. Other studies provide converging evidence on highly accurate pseudoword reading in more regular orthographies than English (Coenen, van Bon, & Schreuder, 1997; Cossu, Gugliotta, & Marshall, 1995; Holopainen, Ahonen, & Lyytinen, 2001; O ¨ ney & Durgunoglu, 1997; Pinheiro, 1995; Porpodas, 1989; Wimmer & Hummer, 1990). Similar results have also been shown by a number of studies where reading development in English and other orthographies have been compared (Ellis & Hooper, 2001; Frith, Wimmer, & Landerl,1998; Goswami et al., 1998; Goswami, Ziegler, Dalton, & Schneider, 2001;Landerl, 2000; Landerl et al., 1997; O ¨ ney & Goldman, 1984; Seymour et al.,2003; Wimmer & Goswami, 1994).
Grunden til at dette er relevant, er at dansk ortografi minder meget om engelsk, og engelsk er vistnok altid den værste i disse undersøgelser. Jf. fra introduktionen:
To date, there have been no comprehensive attempts to quantify and compare the transparency of different orthographies, although some orthographies have been subjected to a computational linguistic analysis. For the English language, 31% of all monosyllabic words have been found to be feedforward inconsistent (in the direction of spelling to pronunciation; Ziegler, Stone, & Jacobs, 1997). The corresponding inconsistency is reported to be 12% in French monosyllabic words (Ziegler, Jacobs, & Stone, 1996), and 16% in German monosyllabic words (J. Ziegler, personal communication, February 20, 2001). It is worthy of note that the above-mentioned consistency calculations are based on spelling body–rime correspondences and not grapheme–phoneme correspondences. Seymour, Aro, and Erskine (2003) have presented a hypothetical classification of European languages according to their orthographic depth at the level of grapheme–phoneme correspondences. Based on the expert opinions of COST A81 representatives, they suggest that, of the orthographies included in the current study, English is the most inconsistent when placed on the continuum of orthographic depth. In degrees of increasing consistency, it is followed by French, Dutch and Swedish, German and Spanish, and Finnish as the most consistent orthography that displays regular and symmetrical grapheme–phoneme correspondences.
Jeg har ikke set en sammenligning af dansk og engelsk endnu. Jeg kunne godt forestille mig, at engelsk er mere uregelmæssig. En indirekte måde at måle det på er, at se på hvor besværligt det er at opnå perfekt fonemisk sammenhæng. Dette er meget svært for engelsk, måske sværere end for dansk.
—-
British Journal of Psychology (1991), 82, 527-537
G. Thorstad
Abstract:
The effect  of   the  regularity of  orthography on  the acquisition of   literacy  skills was studied by comparing the reading and spelling of 70 Italian children aged 6-11 years with that of 90 English children learning traditional orthography (t.o.) and 33 children aged 6-7 years learning the initial teaching alphabet (i.t.a.), using an Italian passage for adults which was also translated  into English. The Italian children  learned to read at an earlier age than the English t.o.  children,  but not than the English i.t.a. children.  The English t.o. and i.t.a. children could read more words  than they could spell, whereas the Italian children could spell most of the words they could read and even some they could not read. The English children  read  fast  and  inaccurately,  whereas  the  Italian children  read slowly and accurately using a systematic,  phonological  strategy until  10 years,  when they read fast and accurately. All the children used a phonological strategy in spelling, but   only   the   Italians  were  mostly   successful.  Thus   the   results   suggest   that,   if   the orthography  is predictable and  invariant,   the children use a systematic,  phonological strategy and learn to read and spell more quickly and accurately.
—-
Scandinavian Journal of Psychology, Volume 46, Issue 3, pages 263–272, July 2005
HOLGER JUUL, BALDUR SIGURDSSON
Abstract:
Spelling of cross-linguistically very similar nonwords was compared in 115 Danish and 77 Icelandic children (primarily 3rd and 4th graders). Danish children made more errors than Icelandic children on word medial consonant doublets and on word initial consonant clusters, even when the groups compared were matched on simpler spelling tasks. These results suggest that the acquisition of phonemic encoding skills is slower in “deep” orthographies such as Danish than in more “transparent” orthographies such as Icelandic. The effect of orthography was expected for consonant doublets because of the relatively more complex sound-letter correspondences in Danish. For consonant clusters, however, sound-letter correspondences are perfectly regular in both languages. The study thus points to the conclusion that even the mastery of regular sound-letter correspondences may be delayed in deep orthographies.
Særligt interessant fordi at det har med dansk specifikt at gøre. I stedet for at kikke på hvordan engelsk klarer sig ifht. andre sprog med gode ortografier.
—-
British Journal of Psychology (2003), 94, 143–174
Philip H. K. Seymour, Mikko Aro, and Jane M. Erskine in collaboration with COST Action A8 network
Abstract:
Several previous studies have suggested that basic decoding skills may develop less effectively in English than in some other European orthographies. The origins of this effect in the early (foundation) phase of reading acquisition are investigated through assessments of letter knowledge, familiar word reading, and simple nonword reading in English and 12 other orthographies. The results conŽ rm that children from a majority of European countries become accurate and fluent in foundation level reading before the end of the Žfirst school year. There are some exceptions, notably in French, Portuguese,
Danish, and, particularly, in English. The effects appear not to be attributable to differences in age of starting or letter knowledge. It is argued that fundamental linguistic
differences in syllabic complexity and orthographic depth are responsible. Syllabic complexity selectively affects decoding, whereas orthographic depth affects both word reading and nonword reading. The rate of development in English is more than twice as slow as in the shallow orthographies. It is hypothesized that the deeper orthographies induce the implementation of a dual (logographic + alphabetic) foundation which takes more than twice as long to establish as the single foundation required for the learning of a shallow orthography.
Se især:
The tests for effects of orthographic depth in the complex syllable series identiŽed Danish and English as the two languages which differed sharply from the others. Again, both the logographic process (word reading) and the alphabetic process (nonword reading) were affected. The Danish Grade 1 results displayed enormous variability , extending from non-readers up to fully competent readers (Fig. 7), and included refusal and word substitution errors. The Grade 1 and 2 groups both showed enlarged lexicality effects, indicating a special difŽculty in developing effective nonword decoding. These outcomes were all present in a much more extreme form in the Scottish results. Mean accuracyin the P1 sample fell below 50% . Some children were unable to read and others had dissociated patterns of word and nonword reading analogous to those reported by Seymour and Evans (1999)—alphabetic dyslexia, in which nonword reading is massively inferior to word reading, and logographic dyslexia in which good nonword reading is combined with poor word reading.
The delayed acquisition of foundation literacy acquisition in Danish and English can be interpreted as a combined effect of syllabic complexity and of orthographic depth. Both languages have a complex syllabic structure and an inconsistent system of
grapheme–phoneme correspondences. The more extreme  effects observed in the Scottish sample could be a product of the relative immaturity of the children (the difference in starting at 5 vs. 7 years) or of the greater inconsistency of the English
orthography.
The use of the regressionmethod made it possible to estimate the amount of reading experience readers of English needed to match European mastery and uency levels. For familiar words (study 2), a BAS reading age in excess of 7 years was necessary . This was also true of simple nonwords (study 3) where reading ages above 7.5 years were needed. The results closelyparallel the earlier Žndings bySeymour and Evans (1999) and Duncan and Seymour (2000). These studies found a strong correlation (> + 0.8) between familiar word reading and BAS reading age, and a slightly weaker one (> + 0.6) between nonword reading and reading age. They also pointed to 7 years as the reading age at which foundation literacy acquisition was normally complete. Given that the BAS scale starts at 5 years, this suggests that readers of English require 2½ or more years of literacy learning to achieve a mastery of familiar word recognition and simple decoding which is approached within the Žrst year of learning in a majority of European languages. Thus, the rate of foundation literacy acquisition is slower by a ratio of about 2.5:1 in English than it is in most European orthographies. We were not able tomake a similar estimate for French or Portuguese. However , the results for Danish suggested that two years may normally be required to achieve mastery of simple decoding (study 3).
Altså, vi bruger 2 år mere end normalt (ifht. andre sprog) på at lære at læse dansk. Dette er meget, meget dyrt samfundsmæssigt.
Den komplekse stavelsesstruktur som tilsyneladende er del af årsagen, kan vi ikke gøre så meget ved, da det ville være dumt at ændre stavemåderne uden at ændre talesproget. Og det er meget svært at bare lige ændre talesproget. Omvendt så kan vi relativt nemt gøre noget ved stavemåderne – nemlig at lave dem om.
—-
Frank Seifart
A  second, more substantial point  to be made here  is  that  from  the perspective  of  psycholinguistics,  “the  optimal  orthography  for  a  beginning reader  is not the same as for a fluent reader” (Dawson 1989: 1). This general  statement derives  from  the  finding  that  advanced  readers heavily  rely on what is called a “sight vocabulary”, i.e. written words are recognized as entire units and processed as such, without breaking them down into units of the  sound  structure. For  that  reason, advanced  readers benefit  from orthographies  that preserve  the graphic  identity of morphemes. A sight vocabulary allows readers to quickly recognize words in written messages without much  specification  of  phonetic  details.  A  high  reading  competence  also allows to make full use of contextual cues, which may require some going back and forth in a written message to disambiguate homographs. Because of  the  relative  importance of a  sight vocabulary and  the  relative unimportance of phonetic detail, advanced readers benefit from deep orthographies rather than shallow ones.
For beginning readers, however, things are different. The acquisition of a deep orthography at first exposure is relatively difficult because the written form may differ  significantly  from  the actual pronunciation and may have to be memorized in a first phase. Compared to these, shallow orthographies, i.e.  orthographies  that  represent  linguistic  forms  in  a way  that  is  close  to their actual pronunciation  in each context, are considerably easier  to  learn for  a  beginning  reader  (and  writer),  including  second  language  learners. Wherever languages display heavy morphophonological processes, orthography developers face the problem of either choosing a shallow orthography for the beginning reader or a deep orthography for the advanced one.
Med andre ord, angiveligvis er morfologiske stavemåder bedst for ‘avancerede læsere’, mens at fonemiske er bedre for alle andre. Alle avancerede læsere starter også som ikke-avancerede – når de er børn.