Introduction
One tricky problem with English spelling is the variation in the nominalization suffix often represented in discourse as ‘shun’, mainly between -tion and -sion. Current ELT textbooks have generally not discussed rules for its spelling. However, following online resources, some basic rules are in current debate, with two main schools of thought, each falling in line with one of two approaches that can be called the ‘word-based model’ and the ‘base-word model’. In this article, I show the base-word model to be preferred, determine the actual suffix and its underlying form, and elaborate on base-word ending clues to yield a general synchronic rule for changing from -tion to -sion, albeit with exceptions.
Background
In modern English studies, an intriguing question frequently recurs: why are English words not always written the way they are spoken? Classic examples include the use of digraphs like < ph > for /f/ and < gm > for /m/ as in ‘diaphragm’ (Achiri–Taboh, Reference Achiri-Taboh2017).
English spelling has been of significant interest to scholars since the 1950s (e.g., West, Reference West1955) and a major problem even to native speakers (West, Reference West1965). One particularly intractable problem concerns the spelling variation of the nominalization suffix often represented in discourse as ‘shun’, underlined here in dictation (with t) and extension/expression (with s). Here, the suffix is pronounced /ʃǝn/ (cf., for example, the Cambridge Advanced Learner's Dictionary [CALD], 3rd edn., 2005: 345, 438–9); but note cases like equation and persuasion with the same suffix rather pronounced /ʒǝn/ (CALD, 2005: 419, 941).
Most shun-words are spelled with -tion or -sion (Harnew, Reference Harnew2007). However, other spellings do exist: viz, -cian (professions: beautician, musician, physician), -sian (origins or nationality: Asian, Caucasian), -tian (also origin: Lilliputian, Venetian (note US dietitian)), and some arbitrary ones like -xion (complexion, connexion, crucifixion), -cion (coercion, suspicion), -shion (cushion, fashion), and -cean (crustacean, ocean). Those that name professions and origins are usually straightforwardly predictable, and the arbitrary ones are fewer and can be learnt by heart. For example, those naming professions are generally based on nouns with -ic(s) endings like magic and politics. As Harnew points out, cushion and fashion are the only two common English words with the ending spelled -shion. So the problem is mainly the -tion/-sion variation.Footnote 1 Thus, I restrict my referents of shun-words to those with either variant.Footnote 2
Following Seely (Reference Seely2005: 210), spelling problems can best be understood in terms of patterns, two of which he identifies. The first has to do with the correspondence between sounds and orthography, in which sounds are often spelled in several different ways across words. The second has to do with the orthographic correspondence between word stems and affixes that may cause variations in the spelling of both. It is with this second pattern that I am concerned.
Current English grammar textbooks have generally not discussed rules for the spelling of shun-words. However, following online resources like Spellzone.com and Dailywritingtips.com, such rules are currently in debate, with two main schools of thought. One considers the spelling of the word as a single whole (as with mission and station), and then looks at its internal structure for clues – we can call this the ‘word-based model’. The other looks at the word as a combination of a base-word/stem plus a suffix (as with dictate/dictation and permit/permission), and then considers the base-word ending for clues – we can call this the ‘base-word model’. In this article, I will show that the base-word model is to be preferred, determining what the actual suffix and its underlying form are, and elaborating on the clues from the base-word endings and exceptions.
Previous literature on ‘shun’
1. Affixation and the rules of English spelling
As Taylor (Reference Taylor1981) points out, English spelling is by no means as unsystematic as many assume it to be. Thus, spelling rules are generally developed and mastered in the writing of English words.Footnote 3 This follows from Pinker's (Reference Pinker1999: 269) Word-and-Rules Theory that holds that the workings of language are essentially two, that is, (1) ‘words in the sense of memorized links between sound and meaning’ and (2) ’ … regular inflection … computed by a mental operation that does not need access to the contents of memory … ’. Deducing from this requires us to either keep to rules or memorize spellings. As indicated earlier, such rules may include those that show how the adding of affixes changes the spelling of the base-word. See Perrin (Reference Perrin1990) and Seely (Reference Seely2005), for example, for details on existing rules.
As it stands, however, affixation also causes a variation in the spelling of the affix, as is the case with the underlined negative prefix in impossible (im - before a bilabial consonant) and insubordinate (in - elsewhere). This is what happens in the case of the shun-suffix. As we shall see, what looks like an ss-form is actually just an s-form preceded by s. One question that might be asked here is whether there can be spelling rules to regulate the spelling of ‘shun’. Spellers may get confused in the spelling of ‘shun’ because they are not readily exposed to rules to follow. What this means is that, with so many derived shun -words, crude memorization is not a suitable option. As Adams–Gordon (Reference Adams–Gordon2010: 12) puts it, ‘the greatest force which impedes the learning of spelling is a lack of interest or presence of undesirable attitudes towards instruction’. Such a force may afflict the learning of spelling once there are no rules that can alleviate the yoke of memorization.
2. Current debates
As pointed out earlier, two main types of rule characterize the debate on shun-words. Following the word-based model, the spelling of the suffix depends on the preceding letter. Thus, as shown on the website Spellzone.com, the suffix generally takes t except after l (compulsion), r (excursion), s (procession), and n (expansion).
With respect to the base-word model, the actual form of the suffix is predicated upon the nature of the base-word ending. Thus, as shown on the websites Howtospell.co.uk and Skillsworkshop.org, the t-form is the commonest form, while the s-form is used when the base-word ends with -d (comprehend/comprehension), -de (collide/collision), -se (infuse/infusion). According to these websites, the s-form has an -ssion variant used when the base-word ends with either -mit (as in admit/admission) or -cede/-ceed (pronounced /siːd/) (concede/concession and succeed/succession), and an -ion variant which is simply added when the base-word ends in -ss (discuss/discussion).
Looking at the two models, the base-word model has more clues to grapple with and therefore looks more cumbersome. In addition, it requires the speller to beware of changes in the base-word. At face value, the word-based model seems more economical, since we can talk of it as simply requiring us to follow the rule described in (1).
(1) Use the t -form everywhere except
a. after l, n, r, and s, and
b. watch out for some exceptions.
However, such a rule takes for granted that the speller already knows how to spell the entire word, except for the form of ‘shun’. Thus, given the status quo, the rule does not seem to be speller-friendly. For example, an important problem is to predict the letter preceding ‘shun’. Specifically, in spelling ‘shun’ in the three-syllable word depression, for example, what informs the speller that the preceding letter is s, and not e which is what is clearly perceived phonologically at the end of the middle syllable? A similar question would be asked in cases like expansion and extension where, preceding ‘shun’, [t] is perceived after [n] (CALD, 2005: 434, 438), and in cases like incursion and perversion where [ɜː], and not [r], is perceived before ‘shun’ (CALD, 2005: 648, 942). It would therefore be more profitable to consider the base-word first. In the rest of this article, I elaborate on the workings of the base-word model, starting with determining what the actual suffix is.
The suffix
It might be thought that the suffix under consideration is actually -ation, -ition or –ion as in Eastwood (Reference Eastwood2002: 370), or that the latter form is a completely different suffix – see Jackson and Amvela (Reference Jackson and Amvela2000: 76), where -ation and -ion are seen as distinct suffixes (also see Harnew, Reference Harnew2007). Huddleston and Pullum (Reference Huddleston and Pullum2005: 284) state that ‘in persuasion we have a change in the consonant at the end of persuade’, and by this, the understanding is that the letter d on the base-word changes to s such that what is actually suffixed is -ion. This view actually seems to gain support from such pairs of words as in Table 1, where the base-words end in double-s.
A support for -ion?
However, this may not be the case for empirical reasons. Firstly, as we will see later, in a case like persuasion, it rather appears the consonant at the end of persuade simply drops when the suffix is added as -sion and pronounced /ʒǝn/. This is easily demonstrated with examples like extension /ɪkstentʃǝn/ and pretension /prɪtentʃǝn/ where the consonant at the end of the base-word, an alveolar stop, clearly stays, at least in pronunciation (CALD, 2005: 438, 998), with motivation from the preceding alveolar nasal (i.e., with the presence of an alveolar -effect), and only changes from [d] to [t] in harmony with the voiceless initial [ʃ] of the suffix. Thus, it can be said that, in permission, the final-t of the base-word changes to s before the suffix. Besides, as we shall also see, in words like reformation, epenthesis takes place between the base-word and the suffix. With respect to the facts in Table 1, it can then be said that the final-s is dropped, and the suffix is added in the s-form motivated by the retained base-s (literally maintaining the double-s). Finally, as far as I can tell, the pronunciation /ʒǝn/ can only be used after a vowel, whereas /ʃǝn/ can be used both after vowels and consonants.Footnote 4 This suggests that, in the examples in Table 1, the suffix is actually placed after the consonant [s], as it is pronounced /ʃǝn/. Thus, writing about the spelling of ‘shun’ on Dailywritingtips.com in Reference Maddox2010, Maeve Maddox is right to point out that ‘“ssion” is not a valid spelling of “shun.” In a word like succession, for example, the first s belongs to the second syllable: suc-ces-sion’. With these arguments, it can be taken for granted that the actual suffix is -sion/-tion. The fact that the suffix exhibits an orthographic variation is therefore clearly intriguing, especially as others like -ment and -less do not.
If the suffix actually begins with either t or s, then to formulate rules of writing it in its correct forms, it would be appropriate to start by saying that one is the base, the other derived. A reasonable guess is that the t-form is the base, not least because a majority of shun-words are spelled with t. Thus, we can postulate that the t-form is generally used, while the s-form is restricted to specific environments. In the next section therefore, I start by looking at where the t-form is more likely to be used.
Forms and clues
1. The t-form
To start with, when the base-word ends with a double consonant as in (2) (see Table 2 below) or a single consonant preceded by a long vowel as in (3), an epenthetic vowel, often [ɪ] or [eɪ], is added before ‘shun’ (often with a vowel change in the final syllable of the base). Exceptions include the word examine which accepts epenthesis to yield examination, even though the final consonant is neither doubled nor preceded by a long vowel.
Adding the suffix
However, many other words harbour similar conditions, but do not accept epenthesis. In some cases where the final consonant of the base-word is preceded by a long vowel, the former is rather replaced with a different consonant (often with a change in the preceding vowel) as in (4) – when the final consonant is a voiced labial or labio-dental, it changes to the voiceless bilabial stop [p]; if it is the voiceless palato-alveolar [s] spelled with c, it changes to the velar stop [k] (although the c-spelling is maintained). Note that the alveolar-effect prevents such substitution in words like detain; in conjoin the preceding diphthong is a complex with a back vowel that changes the alveolar nasal into a velar nasal and introduces the c before the suffix.
What determines the preference for consonant change and the choice of consonant in cases like (4) as opposed to epenthesis and the choice of epenthetic vowel in cases like (2/3) before the addition of the suffix is not clear, and we simply have to get used to the spellings. What is clear, however, is that, once there is epenthesis or consonant change, the t-form is required.
In cases where the base-word simply ends with a t, preceded by a vowel as in (5/6) or another consonant as in (7), the t drops (with any mute-e) and the t-form is added, generally pronounced /ʃǝn/. Observe that the final syllable vowels in the base-words in (5) generally pattern like the epenthetic [eɪ] which triggers the use of the t-form. Also, the derivatives in (7) pattern like those in (4) whose base-words undergo a consonant change for the t-form.
However, exceptions include words like invitation and competition where instead of dropping the final-t of the base-words compete and invite to get ‘competion’ and ‘invition’ (as in (6) with contrite/contrition or ignite/ignition), epenthesis rather takes place like in (2/3); and expectation and suspicion where instead of dropping the final-t of the base-words expect and suspect to get ‘expection’ and ‘suspection’, as in (7), there is also unexpected epenthesis to have expectation and the unexpected suspicion with the suffix spelled -cion without epenthesis. I return to pairs like divert/diversion below.
Finally, there are cases whose base-words end with a t preceded by an alveolar consonant; these do not drop the final-t when the suffix is added (arguably due to the alveolar-effect) as in (8) where it is retained (though only in pronunciation) and (9) where epenthesis rather takes place – some words like indent may exhibit either process as in indention or indentation.
2. The s-form
One general observation here is that when the base-word ends with the consonant [d], spelled with a single d (with/without a mute-e), the s-form is used (see Table 2 above). However the pronunciation of the suffix is not the same on all the words. The spelling of the base-words in this category usually ends either with -nd or -V + de, V for any vowel. When added to -nd base-words, [d] stays due to the alveolar-effect (though only in pronunciation, changing to [t]) and the suffix is pronounced /ʃǝn/ as in (10). Some words may behave the same when their base-forms do not end with -nd, an example being tense/tension. Other exceptions are seen in (11) with -nd base-words which rather take the t-form, accompanied in some cases by epenthesis, and we are again left to try to recognize them individually. When the suffix is added to -V + de words, -de is dropped, and the suffix is pronounced /ʒǝn/ as in (12).
Besides base-words that end with d, a number of syllable endings typically trigger the use of the s-form. Thus, words that end with -fuse, to start with, generally drop the final-s (or [z] sound) together with the mute-e and take the s-form pronounced /ʒǝn/ as in (13) – compare accuse which does not end with –fuse and so rather triggers epenthesis and takes the normal t-form (cf. (3)). There are also the base-words that end with -vert as in (14) which drop the final-t and take the s-form pronounced /ʃǝn/ – compare exert and insert which do not end with -vert and therefore take the normal t-form (cf. (6)).
Furthermore, there are the words that end with -pel as in (15). Here, the final consonant is not dropped, although the vowel in -pel changes to [ʌ]. A word that behaves like the base-words in (15), although its final syllable is not initiated by p, is convulse which yields convulsion. Then there is the set of disyllabic base-words that end with the syllable -C + ise as in (16), C for any consonant. Compared to the examples in (3) earlier, one would expect epenthesis and the use of t as in improvise/improvisation. However, it is the final-se that is dropped and rather the s-form used (pronounced /ʒǝn/) – note that the key here is for the base-word to be a disyllabic word with a -C + ise ending.
Finally, we have two sets of examples as in (17–18) in which the base-words end with the stressed syllables -mit and -cede/-ceed (pronounced /siːd/), respectively. In these cases, the final consonant does not drop like we have seen in other cases; it rather changes to s before the s-form is added (to yield a double-s scenario) – compare edit, exhibit and prohibit which neither end with -mit nor -cede/-ceed and so rather take the regular t-form as in (6). Recall that, when the base-word ends with a double-s, the final-s is dropped and the s-form is used (cf. Table 1), (literally) maintaining the double-s.
3. Summary
In summary, the key facts discussed so far take us to a general spelling rule for adding ‘shun’ following the base-word model as in (19):
(19) Use the t-form on the base-word, except
a. when the word ends with a d(e) [Drop d(e) and add -sion] (cf. (10/12)), or
b. with any of the syllables -fuse, -vert [Drop final se/t and add -sion] (cf. (13/14)), or
c. the syllable -pel, [change [e] to [u] and add -sion] (cf. (15)); or
d. when the word is a disyllabic word that ends with the syllable-C + ise (C = any consonant) [Drop final s(e) and add -sion] (cf. (16)), or
e. a word that ends with the stressed final syllable -mit [change t to s and add -sion] (cf. (17)), or
f. the stressed final syllable -cede/-ceed [maintain a single inter-consonantal e, change d(e) to s and add -sion] (cf. (18)), or
g. when the word ends with a double-s [Drop the final s and add-sion] (cf. Table 1).
Although this general rule includes as many as seven conditions, once spellers get used to the idea of picking out the base-words from complex shun-words and considering their endings, it becomes easy to establish a link between each condition and -sion following the Word-and-Rules Theory.
However, like most other rules of English spelling, the rule is not (actually) based on principles, and therefore, as we have noticed, has exceptions. Openings therefore exist for further investigation. In the meantime, we need to master both rule and exceptions; as seen earlier, some of the exceptions also follow distinct and describable patterns that can be mastered.
Non-derived words
As indicated earlier, ‘shun’ also regularly features as part of simple non-derived words, also typically spelled -tion/-sion; recall that our interest deviates from the few cases that are spelled otherwise. Examples with the t-form include auction, caution, diction, mention, motion, nation, notion, portion, potion, ration, sanction, sedition, and station; those with the s-form include excursion, mansion, occasion (pronounced /ʒǝn/), pension, torsion, version, and vision (also pronounced /ʒǝn/); and those that take double-s include aggression, mission, passion, and session.
The same question arises: where do we use -sion as opposed to -tion? One way of answering this question is to assume that they generally also have to be spelled with the t-form, while checking for patterns similar to those illustrated in Tables 1 and 2 for the use of the s-form, like in (20–21).
(20) The word version patterns like those in (14) whose base-words end with the syllable -vert.
- (21)
a. Words like mission and session pattern like those in (17–18) whose base-words end with the stressed syllable -mit or -cede/-ceed.
b. Words like aggression and passion pattern like those in Table 1 whose base-words end with double-s.
However, many of the words do not seem to fit within specific patterns. For example, in pairs like mention/pension and portion/torsion, ‘shun’ is spelled in the two different forms whereas, in each pair, the preceding syllables end in the same way. Also, the word occasion which clearly patterns like those in (2–3) that trigger epenthesis and would be expected to take the t-form (as in station, nation), rather takes the s-form. Thus, non-derived shun-words seem to be largely unruly, and mainly have to be mastered individually and written the way they have been in the long and convoluted history of English spelling.
To conclude, while it is well documented that the problem of English spelling hacks mainly into linguistic history to the extent that we are left to make do with patterns and rules to help spell more accurately, I hope this article has gone some way to resolving the debate on how to spell ‘shun’.
Acknowledgements
Original motivation to write this article came from a 2014 MA English morpo-syntax classroom discussion; I immensely thank the students of that class, especially Grace Kometa, who stressed the difference in the pronunciation of ‘shun’ in the word equation from other words that end with ‘shun’. Clive Upton has read different versions of the article and provided guidance and thought-provoking comments, and I owe him a huge debt of gratitude. I am also grateful for very helpful comments from two anonymous reviewers, and for final checks and fine-tuning by Nicholas Groom. Finally, I have benefitted from a semestrial research allowance from the Faculty of Arts at the University of Buea, Cameroon, and a quarterly research modernization allowance from the Cameroonian Ministry of Higher Education for which I am grateful.
BLASIUS ACHIRI–TABOH is a senior lecturer of Linguistics at the University of Buea in Cameroon where he teaches Generative Syntax, Semantics, and Bantu and English Morpho-Syntax. His research affiliations lie in Bantu/English Morpho-Syntax, the syntax of the clausal left periphery, and current world use of English. He is the author of ‘A generalized question tag in English’ and (recently) ‘The “ph” of English orthography; a digraph or a sequence of separate phonemes?’ both published in English Today. He is currently exploring further into the theoretical basis of tag questions and the stigmatization of ain't. Email: achiri.blasius@ubuea.cm.
