Patterns of English Consonant Clusters in E. M. Forster’s ‘‘The Road From Colonus’’

A consonant cluster (henceforth CC) is a combination of two or more consonants without a vowel between them. There can be initial, medial and final CCs. Additionally, there are initial CCs with medial or final ones, and medial CCs with final ones. To the best knowledge of the researcher, patterns of English consonant clusters in Forster’s ‘The Road from Colonus’’ has not been handled previously.The present study aims at analyzing CCs in such a short story and finding out patterns of such clusters. Those consonant sequences are statistically investigated in order to show their most prominent patterns and their frequency. Accordingly, it can be hypothesised that medial patterns of CCs are more prominent and commonly used than initial and final ones whereas the frequency of final patterns of CCs is higher than that of initial and medial ones. The results of analyzing CCs in the short story under investigation indicate that there are (31) patterns of two and three – initial CCs whose frequency is (335), (79) patterns of two, three and four – final CCs which are repeated (552) times and (122) patterns of two and three– medial CCs whose frequency is (355). Such results lead to the conclusions that medial patterns of CCs are more than initial and final ones but the frequency of the patterns of final CCs exceeds that of initial and medial ones.


INTRODUCTION
A consonant sequence is mentioned to mean a run of consonants uninterrupted by a vowel or a pause. In this case, CCs in English are classified into initial, medial and final according to their positions within words. Initial CCs, known as onsets, comprise two or three consonants. They are referred to as two -segment initial clusters such as: from /frəm/, dry /drai/, stay /stei/, and three -segment initial clusters such as: street /stri: t/, square /skweə /and spread /spred/. Medial CCs occur in the middle of words between two vowels in a single word which means that the consonants belong to different neighboring syllables. According to (internet 1), this type of CCs is known as inter -syllabic. It can include two, three and four segments. Examples of two -segment medial clusters are'' salted ''/so: ltid/ and ''mended'' /mendid/. Three -segment medial CCs are the ones in ''tempted'' / temptid /and ''glimpses'' /glimpsis/. Four -segment CCs are found in ''exclaim'' /ekskleim/ and ''promptly'' /promptli/. Final CCs, known as codas, occur at the end of words in positions preceded by vowels. They contain two, three and four consonants. Examples of such clusters are ''end'' /end/ and ''ask'' /a: sk/ (two -final CCs). ''next'' /nekst/ and ''text'' / tekst/ (three -final CCs). ''twelfths'  syllable. In a single syllable, onset and coda are optional whereas the nucleus represented by a vowel is mandatory in which case no syllable is formed without a vowel except the syllabic consonants that constitute syllables without vowels. According to Carr (1993: 196), a syllable can be either open or closed. The former means that it ends with a vowel whereas the latter indicates that the syllable ends with one or more consonants. Collins and Mees (2008: 74) classify consonants into five groups as illustrated below: 1. Stops are produced with a completely blocking of the air stream. 2. Fricatives are produced when the air passes through a narrow passage where it causes audible friction. 3. Lateral is produced with a centrally blocking of the air stream, but the air escapes laterally. 4. Nasals are produced when the soft palate is lowered and the air is allowed to pass through the nasal cavity. 5. Approximents are produced when the articulators approach each other.
Such patterns show that both oral and nasal stop consonants are followed by approximants. In this case, only /k/ and /g/ may be followed by four approximants. In addition, some fricatives are followed by some approximants. It should be noted that the fricative consonant /s/ is different from other fricatives in that it is followed by the voiceless stops, nasals such as /m/ and /n/, a voiceless fricative /f/ and approximants.
A single pattern such as /smj-/ as in ''smew'' /smju:/, meaning a name of a bird, can be added to the nine patterns stated above (ibid).

FINAL CONSONANT CLUSTERS
In English, the final CC of a syllable in monosyllabic words comprises two, three or four consonants.

CONSONANT CLUSTERS AND DIGRAPHS
A CC is a sequence of consonants without a vowel between them whereas a digraph is a group of two consonant letters to stand for a single consonant sound. For instance, the word ''ship'' /ʃip/ begins with two consonant letters ''s'' and ''h'' that represent the single sound /ʃ/. On the other hand, the single consonant letter ''x'' stands for two consonant sounds /ks/ as in ''six'' /siks/, /gz/ as in ''exist'' /igzist/, /kʃ/ as in ''sexual'' /sekʃuə/, or /gɜ/ as in ''luxury'' /lʌgɜəri /(internet 2).
It should be noted that the consonant letter ''x'' may produce sounds in two different syllables. Also, there is a combination of digraph as clusters as in: ''length'' /leŋϴ/ with two digraphs ''ng'' and ''th'' representing a cluster of two consonant sounds /ŋϴ/. In the same respect, the word ''lights'' /laits/ ends with four consonant letters ''ghts'' which represent only two consonant sounds /t/ and /s/ which means that the two letters ''gh'' are silent. Additionally, the word ''sightscreen'' /saitskri: n/ has six consonant letters ''ghtscr'' combined together in its medial position to represent a cluster of four consonants /tskr/ (ibid).

SONORITY SEQUENCING PRINCIPLE
According to Fasold and Connor -Linton (2006: 30), sonority is defined as ''relative openers of the vocal tract, which corresponds directly to the relative loudness of a sound''. Low vowels are considered as ''the most sonorous sounds'' where their production involves opening the mouth and that they flow out without an obstruction in the air stream. Voiceless stops are referred to as ''the least sonorous sounds'' because their production requires a complete closure of the mouth where the air stops there for a period of time and then it bursts out followed by explosion. Other sounds are between these two extremes. In this sense, Gussenhoven and Jacobs (2012: 165) affirm that ''the sonority of a syllable increases from the beginning of a syllable onwards, and decreases from the beginning of the peak onwards. Intuitively, sonority is related to the overall acoustic energy of segments''. Sonority scale is stated as follows: ''Obstruents -Nasals -Liquids ([l, r], etc.) -Glides ([w, j], etc.) -vowels''. Accordingly, it can be noted that any initial CCs that reverse the direction of increasing sonority such as [mk-] or [wl-] is less common that one that does not, like [pn-] or [ml-]. Conversely, any rhyme that increases the sonority from left to right like [-lj] is disfavored.
In this case, Carr (1993: 198) states that studying the nature of the syllable thoroughly requires sonority hierarchy. It is indicated that the segments comprising a syllable tend to increase in sonority as proceeding from the edge of the syllable towards the center. Thus, in the word ''priest'' /pri: st/, the /r/ is more sonorous than /p/ and the vowel /i:/ is more sonorous than /r/. When proceeding from the center of the syllable to the final CC, the /s/ is less sonorous than the vowel /i:/ and that /t/ is less sonorous than /s/.
Similarly, Gierut (1999) points out that a sonority sequencing principle is regarded as a universal hierarchy that decides the permissible sequences of consonants within syllables. In this sense, Glements (1990) claims that this principle requires that complex onsets should rise in sonority, and complex codas involve fall in sonority with the center of the syllable being the sonority peak. This principle is followed by most English onset clusters in which case the sonority distance of the first segment in the CC is lower than the second segment in such a CC as in: ''small'' /smo: l/, ''snail'' /sneil/ and ''slow'' /sləu/. However, this principle is violated by clusters beginning with the fricative /s/ plus a voiceless plosive as in: ''spell'' /spel/, ''stone'' /stəun/ and ''scale'' /skeil/.
In the same sense, Collins and Mees (2008: 78) assert that ''sonorants are more likely to feature towards the center of the syllables''. This characteristic is very important for the ordering of consonant sequences. In this case, /kl-/, for instance, in ''clay'' /klei/ is a possible initial CC whereas /lk/ is impossible. On the other hand, the consonant sequence /kl/ cannot occur in final CCs whereas /lk/ as in ''bulk'' /bulk/ is permitted in this position. Gierut (1999) supports this principle by indicating that increasing the sonority distance between the elements in a CC makes the sequence more natural. For instance, the two -initial CC /dr-/ or /tr-/ which moves from lowest to highest sonority is more natural than another cluster such as /sl-/ or /fr-/ that moves from middle to highest sonority. The same thing can be stated about final CCs. For example, the sequence /-rd/ or /-lk/, which has the greatest sonority distance, is more natural than /-rm/ or /-ns/, which has a low sonority distance.

CONSTRAINTS ON CONSONANT CLUSTERS
Each language has certain constraints on the possible consonant sequences. English, for instance, doesn't allow initial CCs including /pn-/, /ps-/or /vw-/ (Collins and Mees, 2008: 74). In English, a number of constraints on syllable structures can be stated as follows: 1. /ղ/, /h/, /j/ and /w/ don't occur in initial CCs. 2. Voiced fricatives such as /v/, /ð/, /z/ and /ɜ/ do not occur as the second element in initial CCs. 3. In three -initial CCs, the first element should be /s/. 4. In initial CCs, /t/, /d/ and /ϴ/ do not combine with /l/. 5. In onsets, nasals do not combine with stops. 6. The combination of nasals with stops in final CCs is homorganic, i.e. the two consonants are pronounced in the same place of articulation. for example, /mp/ and /ŋk/ but not /mk/ and /np/. This rule does not apply to inflected forms such as ''banged'' /bᴂŋd/ and ''rammed'' /rᴂmd/ where they end with the suffix ''ed'' pronounced /d/ (ibid: 75). Kuiper and Allan (1996: 63 -5) assert that some CCs seem difficult to articulate. Thus, combinations such as /spw/, /stl/ and /stw/ are not permitted as initial CCs. Such clusters are different from sequences such as /pdz/, /bfn/ or /gtb/ which are ruled out according to the phonotactic constraints on (ccc) clusters. Additionally, the sequences /gb/, /vŋ/ and /mʤ/ are not permitted in final CCs.
In this sense, Akmajian et al (1997: 92 -3) affirm that combinations such as /bt/, /nk/, gb/, /pt/, /pb/, and /pk/ cannot occur initially though they are permitted internally. However, consonant sequences including /br/, /dr/, /gr/, /bl/, /gl/, /pr/, /tr/, /kr/, /pl/ and /kl/ can be found in onset positions. Native speakers have the ability to identify the well -formed CCs. The difference between allowed and disallowed consonant sequences is that the former consist of a stop followed by a liquid whereas the latter consist of two stops.
In this respect, Gramley and Patzold (1992: 100) assure that sequences such as ''mbit'' */mbit/ or ''dnime'' */dnaim/ are not found initially. However, a nasal and stop can combine in medial position when they occur in different neighboring syllables as in ''bombard'' /bomba: d/ and ''iambic'' /aiaembik/. The letters ''ng'' are pronounced without the stop /g/ when ''ing'' is added to verbs as in: ''reading'' /ri: diŋ/. However, the stop consonant /g/ is pronounced when adding comparative and superlative suffixes to adjectives as in: ''stronger'' /stroŋgə/, younger /jʌŋgə/ and longest /loŋgəst/. On the other hand, such /g/ is not pronounced when ''er'' is added to verbs to form nouns as in: ''singer'' /siŋə/. In medial and final positions, a combination of nasal plus voiceless stops such as /p, t, k/ or the voiced stop /d/ is permitted as in: ''remember'' /rimember/ where the CC /-mb-/ occurs medially and ''sound'' /saund/ where the CC /-nd/ occurs finally, whereas a sequence comprising a nasal plus voiced stops such as /b/ and /g/ can appear only medially as in: ''amber'' /aemb∂/ in which the the CC /-mb-/ occurs in a medial position.

Devoicing
As a matter of fact, nasal consonants such as /m/ and /n/ are voiced but they become voiceless or partially voiced when preceded by /s/ in two -initial CCs as in: ''smoke'' [sm∂uk] and ''snow'' [sn∂u]. In this sense, the lateral, voiced consonant /l/ becomes devoiced when preceded by voiceless plosives /p/ and /k/ in two -initial CCs where the syllable is stressed as in: ''plain'' [plein] and ''claim'' [kleim]. The same thing can be said about the approximants /w/ and /j/ where they become devoiced when preceded by the voiceless plosive /k/ as in: ''queen'' [kwi: n] and ''qute'' [kju: t] (Collins & Mees, 2008: 94). It is worth noting that devoicing is represented by a diacritic such as [0] written under the devoiced consonant.

ANALYSING CONSONANT CLUSTERS IN E. M. FORSTER'S ''THE ROAD FROM COLONUS''
This section tackles the frequency of CCs patterns and their percentages found in the short story under analysis as investigated in the following tables

RESULTS
The results of analysing CCs in E. M. Forster's ''The road from colonus'' tabulated above can be explained as follows:

Two -Initial Consonant Clusters
There are (26) patterns of two -initial CCs mentioned in the short story under analysis. The frequency of such patterns is (237). The pattern /gr-/ is said to be the most prominent one since it is repeated (40) times and its percentage is (16. 87) as shown in Table 1.  Table 2 shows that there are (2) patterns of three -initial CCs whose frequency is (10). These are /spr-/ which occurs seven times and its percentage is (70) and /str-/ that occurs three times and whose percentage is (30). Table 3 comprises (52) patterns of two -final CCs whose frequency is (358). The most prominent of which is the pattern /-nd/ whose frequency and percentage are (41) and (11.42) respectively.

Three -Final Consonant Clusters
There are (17) patterns of three -final CCs whose frequency is (31) as illustrated in Table 4 where /-kst/ is in common use since its frequency and percentage are (6) and (19. 35) respectively. Table 5 shows that there are (3) patterns of four -final CCs. Each pattern occurs once and its percentage is (33.33).

Two -Medial Consonant Clusters
There are (58) patterns of two -medial CCs whose frequency is (165) in the short story under analysis. The most prominent pattern of this type is /-nl-/ since it is repeated (14) times and its percentage is (8.48) as shown in Table 6.

Three -Medial Consonant Clusters
There are (18) patterns of three -medial CCs whose frequency is (40). The most common and prominent pattern of  Table 3. Frequency of two -final consonant clusters    Table 6. Frequency of two -medial consonant clusters this type is /-ldr-/ which is repeated (9) times and its percentage is (22.5) as illustrated in Table 7.

Two -Initial with Final Consonant Clusters
This means that there are two -initial CC in addition to a final one in the same word. There are (43) patterns of this type. The frequency of such patterns is (62). Pattern /st-pt/ is in common use because it occurs seven times and its percentage is (11.29) as in Table 8. Table 9 indicates that there are (12) patterns of two -initial CC with a medial one in the same word. Each one of such patterns occurs once and its percentage is (8.33).   Table 10 shows that there are (40) patterns of two -medial CC in addition to a final one in the same word whose frequency is (68). Pattern /-ms-lf/ is most commonly used since it occurs (9) times and its percentage is (13.23). Table 11 shows that there are (7) patterns of two CC in addition to a final one in the same word. The frequency of such patterns is (8). Pattern /-nd-st-nd/ is in common use since it occurs twice and its percentage is (25).

Three -Medial with Final Consonant Clusters
There are (6) patterns of three -medial CC with a final one in the same word. The frequency of such patterns is (7) as shown in Table 12 where /-ŋgl-nd/ is prominent since it occurs twice and its percentage is (28.57). Table 13 shows that there are (6) patterns of three-initial CC with a final one in the same word. Thefrequency of such patterns is (12). The two patterns /str-ŋϴ/ and /str-gl/ are considered in common use since they occur three times each and whose percentage is (25). Table 14 indicates that there are (23) patterns of two CCs that occur in the medial position of words. The frequency of such patterns is (24). Pattern /-sk-nt-/ is the most prominent because it is repeated twice and its percentage is (8.33).

Miscellaneous consonant clusters
There are certain miscellaneous patterns of CCs as explained below: a. Four -mdial consonant clusters A pattern of four -medial CC such as /-kspl-/ occurs twice in the short story under analysis. b. Four -medial with final consonant cluster There are two patterns of four -medial CC with a final one such as: /-kskl-ʃnz/ and /-kskl-md/.

c. Three groups of two -medial consonant cluster
There is one pattern of this type such as: /-bs-lj-tl-/ which occurs twice in this short story.