User:Apeloza/sandbox

Spelling To Sound Consistency in Reading

Spelling to sound consistency is a phenomenon studied in Cognitive Science in which words that are defined as consistent are easier to access in our memory. Consistency is one of many properties of words (including things like word length and frequency) that changes how quickly and accurately we read. A word which is consistent is one in which the body of a word (also known as the rime) only has one form of pronunciation (example: -at in Hat, Cat, Bat, Mat, etcetera). Similarly, a word is inconsistent if the body of the word has more than one way for it to be pronounced. (e.g. HAVE is not consistent with GAVE or SAVE). This effect appears both in children and adults and appears to be consistent across differing reading skill levels. There are two models that both seek to explain this phenomenon, listed below.

Dual-Route Cascade Model and Parallel Distributed Processing Model

The Dual-Route Cascade Model (shortened to DRC) explains that there are two types of words: irregular words and regular words. Regular words follow something known as grapheme-to-phoneme rules (shortened to GPC), while irregular words do not . From these two types of words, two paths for reading emerge. The first path is the direct route which is what we use when we 'sound out' words, and this is used for irregular words. The second path is the indirect route which is for words that follow GPC rules. In this indirect route the individiual graphemes are turned into phonemes using the GPC rules, in a sense a 'shortcut' which allows for faster processing time.

The Parallel Distributed Processing Model (shortened to PDP) has a different explanation for how we read. The PDP model emphasizes a connectionist approach. Past information influences future system processing for reading. Put simply, each unit of spelling or sound has a connection and those that are most typical are most likely to activate. Rather than being two discrete paths the PDP model represents consistency along a continuum where a word can be highly consistent, highly inconsistent, or somewhere in between. Recent research tends to focus on and agree with the PDP model rather than the more traditional DRC model.

Regardless of the model used to interpret spelling-to-sound consistency, both agree that regular words are simpler for us to recall and consequently faster to read.

Interactions

There are a number of variables that have been found to interact with spelling-to-sound consistency. Frequency (discussed further below) is one of them. Age of acquisition has also been found to interact with spelling-to-sound consistency in such a way that the earlier a word is learned the faster it is processed for both consistent and inconsistent words. There is a debate as to whether imageability (or the ability to visualize a word in picture form in one's head) interacts with spelling-to-sound consistency and this will be covered in more detail in the Controversy section.

High-Frequency and Low-Frequency Words

Spelling-to-sound consistency has been thought to only apply to low-frequency words -- words which are not encountered commonly in one's day to day life. For high-frequency words or words which are encountered commonly in day to day life (example: through) there is little to no change in their recall time. This can be explained through the Dual-Route Cascade model mentioned above: high frequency words with low spelling-to-sound consistency actually travel through the indirect route as they are firmly entrenched in memory and do not need to be 'sounded out'. Through the Parallel Distributed Processing Model this could be explained as the high-frequency words having a large number of exposures which eliminates any corrupting influence similarly spelled (but differently pronounced) words might exert. For low-frequency words this elimination of corrupting influences is not there.

There is some disagreement that spelling-to-sound consistency does not affect high-frequency words. One argument is that high-frequency words are merely not inconsistent or irregular enough to produce an effect. This can be explained through the number of 'friends' a word has (words which are pronounced the same and structured similarly, example: hat, cat). and the number of 'enemies' a word has (words which are pronounced differently and structured similarly, example: HAVE and GAVE). In studies that examine spelling-sound consistency the high-frequency words that are used tend to have a higher friends to enemies ratio than the low-frequency words. It has been found in at least one study that high-frequency words can also exhibit the spelling-to-sound consistency effect if the amount of friends and enemies as well as the types of friends and enemies (example: high-frequency enemies, low-frequency friends) are controlled to be even across both sets of words. However the effect is still larger for low-frequency words. Other studies agree with this notion, highlighting the idea that spelling-sound consistency is not a property of any one word but rather the word as it compares to others like it (or unlike it), which is also emphasized in the PDP model of word processing.

Other Languages

The spelling-to-sound consistency effect can be observed in other languages besides English. Despite differences in phonology and spelling rules the effects can still be seen, though they might manifest differently. In French the shift from spelling to phonology (pronunciation) is fairly consistent (at least more-so than English), but the shift from phonology to spelling is inconsistent. This is a different form of inconsistency that is not examined quite as often in the literature which can be described as feedback inconsistent whereas traditional inconsistency can be described as feedforward inconsistent. In fact, 79.1% of monosyllabic French words are feedback inconsistent with only 12.4% being feedforward inconsistent, contrasted to the English statistics: 76% are feedback inconsistent with 33% being feedforward inconsistent. Furthermore, if one takes all words in French which are considered feedforward consistent, 77.4% of those are actually feedback inconsistent. This highlights the idea that French as a language is not as consistent as first appearances would suggest, and that spelling-to-sound consistency can be observed in French.

German is another language for which spelling-sound consistency has been examined. English and German are similar in structure, but not in their consistency: German is more consistent in pronunciation. Since German is more consistent, there is evidence that native German speakers do not produce the spelling-to-sound consistency effect. A study was conducted which showed that English early readers (age 6) exhibited a consistency effect in word body pairing where the children were able to identify the odd word body out in a word triplet if the body was consistent (roughly 90% accuracy) and had poorer accuracy when the word body was inconsistent (78% accuracy). For German early readers (age 6) there was very little difference between consistent word bodies (68% accuracy) and inconsistent word bodies (64% accuracy). This was probably due to German's more rigid pronunciation rules, mentioned earlier.

The Dutch language exhibits similar effects of spelling-sound consistency as English. While Dutch is similar to German it borrows several words from other language (such as the English computer or the French douche, meaning shower). These words are pronounced differently from the rest of the Dutch language and introduce more inconsistency into Dutch than exists in German. A study was conducted regarding the Dutch language which showed that for these borrowed words (referred to as strange words) which displayed inconsistency effects it was possible to remove the spelling-to-sound inconsistency effect by engaging in regularized reading, a practice where one reads the word as if it was sound-spelling consistent (example: for English, pronouncing HAVE such that it rhymes with SAVE). Whether this effect can be displayed in languages other than Dutch has yet to be explored.

In Neuroscience

Spelling-to-sound consistency has been examined in neuroscience to determine where in the brain the effects of consistency (amongst other variables) might be located in the brain. An experiment was conducted to examine the brain's ability to transform orthographic information into phonological information (that is, turning written word into speech by reading out loud) which spelling-to-sound consistency is intimately related to. Using functional neuroimaging (specifically a PET scan) it has been discovered that sound-to-spelling consistency effects can be observed in the left-front brain region as well as the motor cortex, which suggests a more base reaction to consistency effects rather than spelling-to-sound consistency being purely a higher order phenomenon.

Controversy

There is evidence to suggest that the spelling-to-sound consistency effect can only be observed in words which contain digraphs, which are a pair of letters which form one distinct sound (example: /th/ in English). The experiment found that for both words and nonwords naming displayed a typicality effect only when the words contained digraphs. Those without digraphs did not display a statistically significant consistency effect. The study goes on to note that these particular results are not interpretable by either the Dual-Route Cascade Model or the Parallel Distributed Processing Model. This could have strong implications for the future of spelling-sound consistency research as the majority of it begins with the assumption of either model explaining the general spelling-to-sound consistency phenomenon.

There is also disagreement that English is as inconsistent and chaotic as much of the linguistics literature implies. While it's agreed that English can be presented as chaotic if the focus is on direct representation of phonemes, spelling-to-sound consistency within English is much higher if the position of the phoneme is considered within each syllable as well as the neighbouring phonemes to any one phoneme. It is argued that by introducing this knowledge of the true consistency of English that the face complexity of spelling for new learners can be reduced. This 'hidden consistency' may also have implications for the spelling-to-sound consistency effect in the future.

Lastly, there is disagreement on whether imageability interacts with spelling-to-sound consistency. While some studies have shown an interaction of imageability only with low frequency words on spelling-to-sound consistency such that a low frequency inconsistent word that had high imageability (example: ghost) would be recalled faster than a low frequency inconsistent word that had low imageability (example: suave) , others contest that these experiments failed to properly account for age of acquisition and once age of acquisition is controlled for the effect disappears. This debate is still ongoing with evidence being presented by both sides.

Present and Future

An up and coming area of research into spelling-to-sound consistency involves disyllabic words. Traditionally studies of spelling-to-sound consistency only incorporate monosyllabic words (example: car, late, cup). The issue with conducting research around monosyllabic words is that most of the words we use in our every-day lives are in fact polysyllabic (example: restaurant, toilet, going). Polysyllabic (including disyllabic) words are more complex and thus might change the nature of how we understand spelling-to-sound consistency. One major question around this research is whether a polysyllabic word is examined as a whole word or broken down by its syllables with each syllable pronounced separately.

An experiment was conducted to examine whether disyllabic words specifically were processed in a similar manner to monosyllabic words and whether they also presented a spelling-to-sound consistency effect. It was found that there was a spelling-to-sound consistency effect within disyllabic word processing, particularly around the consistency of vowels within both syllables. This research is the beginning of a new direction within spelling-to-sound consistency research with more complex words (including the possibility of studying polysyllabic words).

Neuroscience is a field that holds much promise for spelling-to-sound consistency moving forward as the majority of research has focused on laboratory experiments which chart performance rather than the actual underlying brain processes. In particular, electrophysiological data which can chart brain processes over time could aid in gaining a further understanding of exactly which areas of the brain work in tandem in reading and in turn produce the spelling-to-sound consistency effect and furthermore how that process changes over time.