Graphic representation of syllables in English

Name
Syllables
Harvey	-hɑr + -vi
Shannon	-ʃæ + -nən
Michael	-maɪ + -kəl
Gertrude	-gɜr + -trud
Sarah	-sɛ + -rə

CHAPTER II. GRAPHIC REPRESENTATION OF SYLLABLES IN ENGLISH AND FUNCTIONAL CHARACTERISTICS OF THE SYLLABLE
There have been a considerable number of studies on monosyllabic word processing in the psycholinguistic literature on reading. It now seems essential to focus on polysyllabic word processing given that monosyllabic words account for a small part of the whole lexicon of a skilled reader. [9.390] This has strong implications for reading models since they are currently dedicated to the processing of monosyllabic words (see Ans et al. 1998, for an exception). With regard to polysyllabic words, the syllable is a major infralexical unit that has received much attention in both the written and the spoken modalities. On the reading issue specifically, studies on the role of syllables have become more and more numerous and the syllabic effects investigated have become more and more subtle. This has engendered the need to use specific tools providing characteristics of syllables. In the first part of this paper, we present data from the literature on the role of syllables in visual word processing and on its relationships with orthographic information. Then we describe the interest of a new tool specifically focused on the syllables of the French language and dedicated to the construction of syllabic materials. In the second part, we present the InfoSyll syllabary, which is a database including numerous variables on the French syllables computed on the words of the lexical database Lexique [10.458]. Finally, we present some quantitative descriptions of the syllabic characteristics of the French language computed from the syllabary.
2.1 Graphic representation of syllables in English
Any syllable as a part of a word has double representation. Its phonetic image is shown in transcription (phonemic or allophonic) and its written notation is shown in spelling (orthography). But the problem is that parts of phonetic and orthographic syllables do not always coincide (let-ter ['let-’ǩ], sin-ging ['sıŋ-’ıŋ). [13.298]
Syllable division in writing may follow the rules which have already been stated above. Then the division of the syllabic structure in orthography is made according to phonological principles (fa-mi-ly, re-gu-lar). Still such a division is not always possible. For example, in the words body ['b$d-’ı], money ['mÙn-’ı], the consonant letter representing the point of syllable division should be added to the next vowel letter in order to escape notation gaps: bo-dy, mo-ney.
Syllable division in writing may be also based on the morphological principle. In this case prefixal and suffixal morphemes are divided from the root one (ir-regul-ar), no matter whether they belong to the same phonetic syllable or not. Graphic representation of syllables in orthography is relevant for language learning only when it is necessary to carry some of the letters over to the next line. This process is usually called syllable separation. [14.593] Special attention is necessary in order to exercise it correctly. There are the following rules:
— a word can be separated only if it consists of more than 5 letters and contains more than one syllable;
— the number of separated letters should be more than one;
— the parts of a word subjected to separation are derivational morphemes used in word-building, but not in flexional ones used in word-changing (be-come, friend-ship, commit-ment, sports-man, volley-ball);
— the suffix -ing can be separated with the preceding consonant if there is a consonant cluster before it: hand-ling;
— suffixes consisting of two syllables can’t be broken and should be separated as a whole: vulner-able;
— suffixes consisting of two letters can’t be separated with the exeption of -ly: surprised, teacher, graphic, but: correct-ly.
Syllabic Effects and Orthographic Information
Examining the role of syllables in reading differs substantially from examining their role in speech processing. Though syllables are phonological units, their orthographic features are also processed when displayed visually. This has led to examining the influence of orthographic information carried by words and syllables in syllabic effects and the relationships between syllabic and orthographic units. This issue was first discussed by Taft (1979) in the English language. He argued that syllables are involved in visual word processing, but that the syllabification of words is based upon orthographic principles rather than pronunciation. [16,512] This proposal was based on the strong ambisyllabicity of the English language, and on the cases of incompatibility between syllabification following phonological principles and the morphemic structure of words (e.g., the word leaking has a morphemic boundary after the letter k while the syllable boundary is after the letter a). Hence, Taft (1979) proposed that the basic orthographic syllabic structure (BOSS) is a functional unit of polysyllabic word processing, the BOSS referring to the first syllable of a word plus as many consonants following its first vowels while not disrupting the 123 J Psycholinguist Res morphemic structure (e.g., the BOSS of leaking is leak) (e.g., Rouibah and Taft 2001; Taft 1979, 2001; but see Lima and Pollatsek 1983, in lexical decision tasks). From the point of view of orthographic regularity, Seidenberg (1987) used the illusory conjunction task (i.e., to detect the colour of a letter in a briefly-presented word displayed in two colours) to assess to what extent the syllabic effects in English were subsumed by statistical orthographic regularities. More precisely, given that the bigram that straddles the syllabic boundary of words is usually less frequent than those that precede and follow this boundary, this would produce an orthographic cue to isolate perceptual units such as syllables. This particular pattern of bigram frequencies, referred to as a bigram trough, would thus be responsible for the emergence of syllabic effects. Additional data in French and Spanish actually suggest that the frequency of orthographic units at the syllable boundary can influence the strength of syllabic effects. Syllabic effects were obtained for words both with and without a bigram trough but the effects were stronger when the bigrams made salient the syllables of the words, that is, when there was a bigram trough at the syllable boundary rather than not (illusory conjunction task: Doignon and Zagar 2005; lexical decision task: Carreiras et al. 1993; but see Conrad et al. 2009). In line with these data, Mathey et al. (2006) showed that the frequency of the first bigram of French words influences the strength of the syllabic frequency effects in the lexical decision task, low-frequency bigrams yielding facilitatory syllable frequency effects while high-frequency bigrams yielded inhibitory syllable frequency effects. Another approach to the relationships between syllabic effects and orthographic information lies in investigating the orthographic information contained by the syllables themselves rather than by the words. [18.245] This leads to the important distinction between the phonological features and the orthographic features of the syllables, and thus to distinguishing between syllable units (phonological syllables) and the written transcription of these units (orthographic syllables). For example, the syllable /si/ can be spelled either si or ci(sirène vs. citron) and in Spanish (ciclón vs. sitiar). The question of the role of orthographic syllables in reading has emerged very recently in the literature on the syllable issue. Studies using the primed lexical decision task have shown that when the prime shared the first phonological syllable with the target word but not the orthographic transcription of this syllable, a syllabic priming effect was obtained (e.g., Alvarez et al. 2004). Applied to syllabic neighborhood, the distinction between orthographic and phonological syllables leads to the distinction between phonological syllabic neighborhood (i.e., whole words sharing a phonological syllable, like the syllable /fi/ in ficelle, physique, philosophy,) and orthographic syllabic neighborhood (i.e., whole words sharing an orthographic syllable, like the syllable /fi/ spelled fi in filer, ficelle, fillette, ficher). When the effects of syllable frequency were separately examined according to the syllable features (Conrad et al. 2007), data yielded an inhibitory effect of phonological syllabic neighborhood (orthographic syllabic neighborhood held constant) and no effect of orthographic syllabic neighborhood (phonological syllabic neighborhood held constant). Furthermore, when the phonological syllabic neighborhood and the orthographic syllabic neighborhood were jointly manipulated, the inhibitory effect of syllable frequency was obtained only for low-frequency orthographic syllables in adults, while the reverse pattern was found in fifth graders (Chetail and Mathey 2009b). These data suggest that the orthographic transcriptions of syllables is an important variable and that orthographic information carried by syllables should be taken into account when studying syllable effects, at least. [20.337] The distinction between phonological and orthographic syllables is particularly relevant in languages with inconsistent phonology to spelling correspondences, such as French or English (see Ziegler et al. 1996). The question of consistency between phonology and orthography can be examined on sound-to-print relationships (feedback consistency: number of 123 J Psycholinguist Res ways to spell a pronunciation pattern) and on print-to-sound relationships (feedforward consistency: number of ways to pronounce a spelling pattern). In English specifically, these two measures of consistency are not symmetric, feedforward consistency being much higher than feedback consistency. This was found at different intra-syllabic levels. When considering the rime of monosyllabic words, 87.6% are feedforward consistent while only 20.9% of words are feedback consistent (Ziegler et al. 1996). Similarly on the onset, nucleus, and coda units, words are more feedforward consistent (respectively 77.30, 35.95, and 61.13%) than feedback consistent (respectively 52.02, 1.22, and 14.45%) (Peereman and Content 1999a). This means that in French, the pronunciation of intra-syllabic units in written words is somewhat ambiguous while such pronounced units lead to various spellings. As underlined by Peeerman and Content (1999b), the consistency between prints and sounds can be analyzed for units at different levels of word structure. This seems particularly relevant when examining the role played by a given unit in a given language. In English for example, correspondences between phonology and orthography are more consistent at rime level than at a smaller level, supporting the idea that rimes are important units of word processing in this language (Treiman et al. 1995). In French, given that syllables are more relevant units than rimes, a natural question concerns the degree of consistency at the syllabic level compared to smaller levels. However, no answer could be provided until now given that correspondences between oral syllables and their spellings were not available.
2.2. Functional characteristics of the syllable
The syllable is a phonological unit that performs the following closely connected functions: constitutive, distinctive and identificatory.
1. The constitutive function of the syllable lies in its ability to be a word or a part of it. It this respect the syllable exercises the connection of smaller and greater language units. On the one hand, it represents the correlation of the distinctive and acoustic features of phonemes. On the other hand, it realizes the prosodic characteristics of speech within the stress pattern of words, the rhythmic and intonation structures of utterances. Thus the syllable sums up specific minimal features of both segmental and suprasegmental levels.
2. The distinctive function of the syllable lies in its ability to differentiate words and word-forms taken separately or in combinations. This statement can be illustrated with the following distinctive oppositions: nitrate [naı-'treıt] vs. night-rate [naıt-'reıt]; lightening ['laı-tn-ıŋ] vs. lightning ['laıt-nıŋ]. In these minimal pairs syllable division changes the allophonic contents of words and thus helps to distinguish between them. The similar distinction is found within language units of greater magnitude: an aim [ǩn 'eım] vs. a name [ǩ 'neım]; we loan [wı 'lǩun] vs. we’ll own [wıl 'ǩun]. These oppositional pairs present differentiation of syllables concerning word combinations [ 22.351]. Sometimes syllable division may even be the basic ground for sentence differentiation: I saw her rise [aı 'so: hǩ 'raız] vs. I saw her eyes [aı 'so: hǩr 'aız]; I saw the meat [aı 'so: dǩ 'mi:t] vs. I saw them eat [aı 'so: dǩm 'i:t].
3. The identificatory function of the syllable is conditioned by the pro-
nunciation of the speaker. The listener’s ability to perceive and identify the exact meaning of a word or a combination of words depends on the speaker’s ability to establish the correct syllabic boundary: pea stalks ['pi: 'sto:ks] vs. peace talks ['pi:s 'to:ks]; my train [maı 'treın] vs. might rain [maıt 'reın]. The realization of the distinctive and identificatory functions of the syllable is closely connected with the notion of juncture, kept by the speaker and taken in by the listener.
Close juncture (conjuncture) occurs between the sounds of the same syllable. Open juncture (disjuncture) occurs between the sounds of two different syllables. Some linguists state that word juncture should be marked in a phonetic transcription with [+]. In this case the differentiation between the oppositional pairs will look as follows: