Ann. of Dyslexia DOI 10.1007/s11881-017-0150-x Elegant grapheme-phoneme correspondence: a periodic chart and singularity generalization unify decoding Louis Gates 1 Received: 20 January 2017 /Accepted: 7 September 2017 # The International Dyslexia Association 2017 Abstract The accompanying article introduces highly transparent grapheme-phoneme rela- tionships embodied within a Periodic table of decoding cells, which arguably presents the quintessential transparent decoding elements. The study then folds these cells into one highly transparent but simply stated singularity generalization—this generalization unifies the decoding cells (97% transparency). Deeper, the periodic table and singularity generalization together highlight the connectivity of the periodic cells. Moreover, these interrelated cells, coupled with the singularity generalization, clarify teaching targets and enable efficient learning of the letter-sound code. This singularity generalization, in turn, serves as a model for creating unified but easily stated subordinate generalizations for any one of the transparent cells or groups of cells shown within the tables. The article then expands the periodic cells into two tables of teacher-ready sample word lists—one table includes sample words for the basic and phonogram vowel cells, and the other table embraces word samples for the transparent consonant cells. The paper concludes with suggestions for teaching the cellular transparency embedded within reoccurring isolated words and running text to promote decoding automa- ticity of the periodic cells. Keywords Automaticity. Decoding . Grapheme-phoneme correspondence . Letter-sound relationships . Phonics Randall, a Harvard theoretical physicist, wrote, BData and theoretical consistency together are the uncompromising arbiters of what is right^ (2015, p. 371). According to this standard, both the letter-sound data and theoretical postulates as stated within grapheme-phoneme generaliza- tions display countless inconsistencies. Summarizing a review of letter-sound studies, Johnston wrote, BIt is impossible to neatly categorize sounds and letter combinations in such a way that simple generalizations will work reliably^ (2001, p. 142). Echoing this idea, Duke declared, BWe can’t boil down English orthography into a few simple generalizations…^ (2014). * Louis Gates [email protected] 1 Columbia School District #400, Burbank, WA, USA L. Gates Common sense and general agreement with Johnston and Duke led the present author to deduce that a renewed study would merit little efficacy. Regardless, letter-sound irregularities became painfully personal when, as a rookie teacher, I failed to help Barry unmask letter-sound patterns—decoding befuddled him like deciphering fish from ghoti. Barry, a bright middle school student with special reading needs, faced the challenge akin to Clymer’s seminal study of letter-sound statements in which he mused that many letter-sound generalizations left him with Bno clear indication as to what was to be done^ (1963, 1996,p.183). Remembering Barry and with a nod to the model periodic chart from chemistry and the singularity theory from physics, the author conducted the following study that improbably led to a decoding table, an associated singularity generalization, and companion teacher-friendly tables that include model words for each transparent letter or letter combination. Against logical odds, the study revealed that shrouded beneath the confusing surface rests a surpris- ingly simple but amazingly cohesive decoding elegance. Pivotal decoding research The letter-sound confusion in the New World began as the mayflower moored with eager young children on board clutching the earliest decoding system embedded within horn- books. These hornbooks typically included a wooden or metal paddle overlaid with a transparent cow horn jacket that protected artful calligraphy on parchment which intro- duced the upper and lower case alphabet, followed by nonsense two-letter phonograms and then a benediction. Predating the age of modern science, phonograms, like the decoding generalizations that would follow, understandably evolved without the aid of grapheme-phoneme research. Basal readers from the late nineteenth through the early twentieth centuries evolved to present up to 18 weeks of nonsense phonograms—ba, ca, da—that Betts described as the Bhiss and groan^ method of teaching reading (1955). Intrigued by their pervasive popularity within popular basal readers, Gates amassed 1200 phonograms from which he identified 167 that occurred with enough regularity to warrant additional study. Gates concluded that even these selected phonograms revealed Ba very complex situation^ (1928, p. 146). Although his powerful pen blunted their teaching, phonograms persist to this day (Kress & Fry, 2015). Primers, which included the Protestant Tutor from England and later Webster’s American primer, known endearingly as the Blue-backed Speller, added phonic general- izations to the single letters and phonograms from the hornbooks. In the 1930s, embrac- ing decoding generalizations eclipsed the declining popularity of phonograms. Brown (1930), for example, popularized the unbridled adoption but notably inefficient general- ization that some schoolchildren sing today as, BWhen two vowels go a-walking, the first one does the talking.^ Curious about their efficacy, Clymer culled 150 generalizations from four popular basal readers. Musing that many of these generalizations left him with Bno clear indication as to what was to be done,^ heparedthemdownto45promising statements. Clymer discovered that 25 of the promising ones, including the two- vowel rule, lacked a minimum transparency of 75%. Furthermore, the remaining 20 transparent generalizations displayed a haphazard array of statements (1996/1963). While flourishing in many early reading classrooms, decoding generalizations and Elegant Grapheme-phoneme Correspondence: A Periodic Chart and... the single letters and phonograms that they reference continue to display wide-ranging confusion (Meese, 2016). Parameters of the study To address the letter-sound confusion, the present study began by creating in Excel a column for the 16,928 words that occurred at least once per million running words within the word study by Zeno, Ivens, Millard, and Duvvuri (1995), exclusive of contracted, abbreviated, dialectical, hyphenated, slang, or proper nouns. The second column matched each word with its pronunciation according to the American Heritage Dictionary (online version); as opposed to some dictionaries, this one negated the need to merge allophones into phonemes and the subsequent criticism that this could draw (review this concern in a pivotal study by Hanna, Hanna, Hodges, and Rudorf (1966), 13–14). The juxtaposition of the words with their dictionary respellings enabled sorting of either the word or the pronunciation column. The limitations of the computer analysis follow. Word sorting The replace and sorting options in Excel helped to classify each periodic cell within the 16,928 words. This included a straightforward classification for some of the cells, such as those within words containing the single consonant z (zip), words that included the ee vowel digraph (see), and words with the ight phonogram (night). Other periodic cells presented more complexity and required multilayered sorting. These included, for example, words containing the single vowel a. In this instance, the Excel color-coding option under the replace feature helped to build a file for all of the words containing the letter a within the Zeno word corpus. Next, to catalog single vowel a words, including words that share this pattern and the letter a tied to other periodic cells, the Excel replace option assisted in applying a simple and interchangeable alphanumeric code (A = 01, B = 02, C = 03;... X = 24, Y = 25, Z = 26). This numeric code included the letter a as a part of a vowel digraph (ail=01.09.L), the final -aCE pattern (bate =B.01.20.05), and the phonogram all (pall =P.01.12.12). The final resorting formed a list of words comprised of the single vowel a cell (at, cat, catch), including words containing this cell coupled with the newly formed code (avail=AV.01.09 L, abate = AB.01.20.05, appall = APP.01.12.12). Furthermore, each decoding cell, whether straightforward or more complex, included at least two blind sorts. Discrepancies between the word sorts of a given cell led to further analysis and sorting of the words to reconcile differences. Transparency standard The study parameters included at least the 75% transparency threshold proposed by Clymer (1996, 1963) but sought 90% transparency. The study limited the analysis to the initial occurrence of an identical letter or letter combination within a particular word—this analysis included just the first of the three a’sandthefirstn in banana. Nonetheless, the research addresses variant letters and letter patterns within individual words—the report includes data, for example, for both the basic single vowel i and the i in the phonogram ight (midnight). In addition, the study considered look-alike letter combinations as if they were the decoding cells L. Gates under study—the consonant letters t and h in juxtaposition within the compound word boathouse counted against the transparency of the th digraph. Vowel and consonant category parameters The parameters for the study included a set of broad parameters for the vowel and consonant categories, which follow in bulleted form for ease of reading: