3.1 The apostrophe in the regular past tense

The adoption of the apostrophe in the inflectional morpheme -ed, as found in the past tense and past participle forms of weak verbs, had a phonological rationale. According to Görlach (Reference Görlach1991: 58), the symbol would be used in those items where pronunciation was not syllabic, but consonantal, such as lov'd. This suggests that the apostrophe would not occur in the past forms of verbs with infinitives ending in -t and -d, such as want and need, since the graph that it omits would be phonologically realised as the syllable /ɪd/. This assumption is corroborated in ARCHER, where there are only six tokens of the like, such as relat'd, dissapoint'd and spirit'd. In the light of the unavailability of the apostrophe in these contexts, these verbs have been excluded from the study altogether. Other than -ed and -’d, the analysis examines the selection of -t and -’t, not only due to their concurrence during the Early Modern English period (Osselton Reference Osselton, Blake and Jones1984: 133‒4; Cook Reference Cook2004: 171‒2), but also to their combination with the apostrophe.

Table 1 presents the distribution of -ed versus -’d over time, on the one hand, and of -t vs -’t, on the other. Overall, the percentages show that the full forms ‒ i.e. the variants without apostrophe ‒ are consistently more frequent than their counterparts.Footnote ⁹ From the mid eighteenth century onwards, the former occur in roughly 95 per cent of the tokens of the past morpheme, leaving the use of the apostrophe in a marginal position. Though these trends in usage seem relevant in understanding the process of standardisation, the normalised frequencies shed yet more light on the orthographic configuration of the morpheme. The variant -ed is, by far, the topmost frequent rendering of the past from the onset of the period and it is followed, at a distance, by the abbreviated form, -’d. In the meantime, -t and -’t are sporadically used, especially since their selection is dependent on the phonological features of the verb in question and is not always systematic. If the infinitive ends with a voiceless consonant, the morpheme is realised as /t/, as in published, and its orthography might as well reflect this sound. However, this criterion is not consistently fulfilled and, as reflected in the data, most verbs are spelled -(’)d by default.Footnote ¹⁰ Moreover, the forms -t and -’t start declining in the seventeenth century and virtually disappear by the nineteenth, a development that Cook (Reference Cook2004: 172) has already traced in drama.

Table 1. The variants -ed, -’d, -t and -’t in the expression of the past over time

More compelling in the present discussion is the distribution of -ed and -’d over time. Although the overall dominance of -ed has already been elucidated in table 1, figure 1 best depicts the ups and downs in the development of these orthographic variants. The figure shows that, despite the overall stability of -ed in the first half of the period, -’d had nearly doubled in frequency in the first half of the eighteenth century. Nevertheless, the apostrophe would quickly lose ground during the next fifty years, presumably dropping to pre-1600s levels. In the meantime, the full form underwent an increase in frequency diametrically opposed to the loss of -’d, the patterns in the figure diverging from each other in a parallel fashion. The nineteenth century witnesses, once again, the stabilisation of -ed, with slight increases throughout, whereas its counterpart continues declining and nearly becomes extinct. The figure also points to the apparent correlation between the decline of -’d and the rise of -ed, since it seems as if the spelling variants were in complementary distribution. Nonetheless, the qualitative analysis of the data shows that the preference for -’d does not correspond to any sort of selectional criterion: the topmost frequent fifty verbs partaking of the apostrophe are not only spelled with -ed as well, but they are also more frequent in the latter case. This demonstrates that both -ed and -’d shared an ecological niche and that the only issue preventing linguistic competition was the very incidence of the full form in the expression of the inflectional morpheme.

Figure 1. Distribution of -ed and -’d in the expression of the past over time (n.f.)

Genre variation has often been considered a decisive factor in the selection of the apostrophe in the regular past (see Görlach Reference Görlach1991; Cook Reference Cook2004; Tieken-Boon van Ostade Reference Tieken-Boon van Ostade and Mugglestone2006; Moore Reference Moore2012). In its most elementary function, as in the case at hand, the apostrophe was ‘purely phonetic [in] marking the omission of a sound, and [it] was therefore frequently used in drama texts’ (Görlach Reference Görlach1991: 58), written to be recited. Similar assertions have been made about other text-types closely related to speech, such as poetry (Moore Reference Moore2012: 800) and private writing, where -’d is said to have ‘lingered much longer’ (Tieken-Boon van Ostade Reference Tieken-Boon van Ostade and Mugglestone2006: 256).Footnote ¹¹ Table 2 presents the distribution of the apostrophe across genres in the seventeenth and the eighteenth centuries to show whether, at least in the period when this type of apostrophe reached its peak, this symbol was more prone to occur in speech-based text types. Interestingly enough, the period 1600–49 sees the rise and development of this kind of apostrophe in English writings and, already at this incipient stage, it amounts to 56.23 occurrences in drama in sharp contrast with the other text types represented early on in the corpus, where it is negligible. The period 1650–99, on the other hand, confirms this same tendency where dramatic texts again reach the highest distribution with 71.79 occurrences, with the exception of fiction, which amounts to 107.3 occurrences.Footnote ¹² It is significant to note that the phenomenon also spreads to other genres like medicine and science and, to a lesser extent, to letters and news. This use of the apostrophe reaches its zenith in the first half of the eighteenth century, when it is also frequent in drama along with other genres such as news (126.94 occurrences), journals (124.82 occurrences) and science (107.21 occurrences), a fact which confirms the diffusion of the phenomenon across the different text types. The period 1750–99, however, shows the systematic decline of this use of the apostrophe across all genres with the exception of drama, where it is found to be frequent, with 42.27 occurrences.

Table 2. The apostrophe in the past tense across genres (n.f.)

Drama, as shown, stands out as the only genre where the apostrophe is found to have a consistent distribution over time, with a significant role since the seventeenth century. Figure 2 presents the classification of the regular past tense forms in the dramatic texts of the corpus both with and without the apostrophe to discern if there was any actual competition with the full suffixed form -ed. The form -’d, on the one hand, has a growing diffusion until reaching its climax at the end of the seventeenth century and, more importantly, it comes to outnumber the full suffixed form -ed in the period 1650–99, perhaps as the immediate result of the phonetic virtues of this symbol in particular environments. The period 1700–49, however, marks the progressive decline of the apostrophe in these texts to the extent that it becomes practically negligible towards the end of the century. This change of attitude was likely to have been the result of the eventual standardisation of the suffixed form at the turn of the century.

Figure 2. Distribution of -ed and -’d in dramatic texts over time (n.f.)

The frequency of the regular past tense apostrophe in those genres written to be read aloud suggests that its phonological rationale ‒ i.e. marking consonantal pronunciation, instead of syllabic ‒ must have been evident among language users. However, this function was never overtly codified, since grammarians seemed to systematically disregard this use of the apostrophe. More particularly, instances of -’d are only ever found in Jonson's and Baker's works, the former including this variant among other ways to shorten conjugated verbs (Waite Reference Waite1909: 94), the latter criticising the practice of eliminating one of the geminate consonants in verbs like stunned – stun'd (Baker Reference Baker1770: 26-7). Nevertheless, these authors never assess the propriety of the construction and neither do others like Lowth (Reference Lowth1763) or Ash (Reference Ash1791). Although this lack of prescription may be understood as implicit acceptance, the apostrophe virtually disappeared in this domain during the nineteenth century, with grammarians doing nothing to either encourage or discourage this development. Instead, the standardisation of -ed seems to have been almost unanimously endorsed by eighteenth-century printers as a whole, a practice that the data in table 2 confirmed. By the end of the eighteenth century, the genres with a higher incidence of -’d were diaries, news and letters, while this variant had virtually disappeared in most of printing.

3.2 The apostrophe in the possessive case

In the synthetic expression of possession, the apostrophe seems to entail the omission of linguistic material. Whereas -'s was formerly believed by some scholars to be an abbreviation of the phrase the boy his car (see Sklar Reference Sklar1976: 178), it is more likely a relic of the case system, a remnant of the genitive inflectional morpheme which, in Old and early Middle English, was realised as -es in masculine and neuter strong nouns, spreading later to all nominals (Mossé Reference Mossé1952: 48‒9; Nevalainen Reference Nevalainen2006: 75). Analogously to the past tense, the apostrophe in the Saxon genitive marks the historical elision of <e>, which in most cases was no longer pronounced. As in present-day English, inflectional -s and -'s only produced the syllabic pronounciation /ɪz/ if appended to word-final sibilants, as in the houses or the mouse's tail. Strang (Reference Strang1970: 109‒10) highlights that, phonologically, the distinction between nominative and genitive singular and plural nouns is realised in a ‘two-term system of contrast’, whereas in writing the system is made up of four terms. On the one hand, nominative singulars remain unmarked while nominative plurals and genitives are systematically marked with a sibilant in speech. On the other hand, whereas nominatives have no apostrophes whatsoever in standard written English, the position of this symbol varies in genitive singulars, the boy's car, and in genitive plurals, the boys’ car. Today, the sole function of the genitive apostrophe is to avoid confusion in written language. However, there is no agreement as to when it acquired such a role.

The seventeenth century was, by most accounts, crucial in the development of the phenomenon. Nevalainen (Reference Nevalainen2006: 74) argues that the apostrophe was introduced in the late 1600s, whereas Görlach (Reference Görlach1991: 58‒9) and Cowie (Reference Cowie2012: 604) claim that by that time the symbol was already widespread and even obligatory. Due to such lack of consensus and for the purpose of outlining the timeline of the possessive apostrophe, table 3 presents the frequencies of -'s and -(e)s for the period 1570–1670 in EEBO. The data show that the forms compete against each other between 1620 and 1640, until the apostrophe becomes dominant in the 1650s and continues growing thereafter. The data in ARCHER, as illustrated in figure 3, also reflect the early development on the part of the genitive apostrophe, in the period 1600–49. Although the structure seems to decrease in frequency throughout the seventeenth century, the -'s continues to be dominant and, soon after, it starts gaining ground at the expense of its counterpart, which dropped sharply and had virtually disappeared by the nineteenth century.

Table 3. The forms -(e)s and -'s in the expression of the genitive over time

Figure 3. Distribution of -(e)s and -'s over time (n.f.)

Despite the early occurrences, the possessive apostrophe can be said to have been introduced into the language more systematically during the late sixteenth century. After a short-lived period of competition, -'s became the dominant form in the 1650s and, in the light of its expansion, it might have been considered standard ‒ or obligatory ‒ by the first half of the eighteenth century. The apostrophe thus seems to have become the marker of the genitive case earlier than most scholarly accounts reckoned, especially when considering the data from a qualitative stance. Indeed, as already mentioned, the genitive singular and the genitive plural have typically been discussed separately, with the latter having undergone standardisation at a later time, namely, the late eighteenth century (Görlach Reference Görlach1991: 58‒9; Cowie Reference Cowie2012: 604). It is highly likely that plural genitives were not as quick to regularise due to their lower incidence in use. Comparing the occurrence of -'s and -s’, as in table 4, proves that the latter amount to a very marginal proportion of all the genitives analysed, representing at best 6.82 per cent of the tokens in the nineteenth century, i.e. only after this form had presumably undergone standardisation.

Table 4. The singular (-'s) and plural (-s’) forms of the Saxon genitive (%)

Regardless of the diffusion and standardisation of the genitive apostrophe over time, prescriptivist grammarians still grappled with the uses of this symbol. In his review of the English grammar tradition, Leonard (Reference Leonard1962: 193‒4) argues that the apostrophe was not very frequently mentioned nor employed in the seventeenth century, whereas it seems to have been ‘fairly well-established in the course of the eighteenth’. Although in his 1640 grammar Jonson (Waite Reference Waite1909: 86) does not include it among his paradigms, where there is syncretism between the nominative plural and the genitive forms, Lowth (Reference Lowth1763: 26) and Ash (Reference Ash1791: 30) discuss the apostrophe as conventional usage though still highlighting its inadequacy in examples such as Thomas's book and fox's, where the elision is orthographic but not phonological.Footnote ¹³ The disapproval of the apostrophe, however, is yet more fervent in the plural. Over time, grammarians put forward a number of proposals for the expression of the plural genitive, which included syntactic and orthographic changes, such as discarding the synthetic structure in favour of the of-phrases, as in house of lords instead of lords’ house (Johnson 1755: n.p., as cited in Leonard Reference Leonard1962: 195) and cutting off one of the inflectional -s terminations (Greenwood Reference Greenwood1753: 65) or reversing the apostrophe as in the warrior‘s arms (Buchanan Reference Buchanan1767: 124‒5). The latter, though closer to the contemporary configuration, were criticised on the grounds of their ‘stiffness and formality’, since they might be ‘of service to the reader only, and not the hearer’ (Mennye Reference Mennye1785: 76). Nevertheless, all this confusion and controversy did not stem from actual usage. Priestly (Reference Priestly1769: 7), who has been described as ‘the most … consistent follower of usage in this period’ (Leonard Reference Leonard1962: 315), briefly explains that plural genitives are expressed by simply adding the apostrophe, as in Stationers’ arms, thereby settling the matter. In any event, the grammarians’ disagreement over the use of the apostrophe, along with the data discussed above, ultimately confirm that these linguistic commentaries had no impact on the language whatsoever.

For the purpose of ascertaining the potential impact of genre variation in the diffusion of the genitive apostrophe, table 5 represents the distribution of the possessive case both with and without the symbol across text types in the period 1600–1799. As shown, the general tendency is the systematic adoption of the apostrophe across all genres in all the subperiods, even though the presence of the -s inflection in the period 1650–99 is still found to outnumber the apostrophe in formal writing, such as news (25.87 vs 9.15 occurrences) and, to a lesser extent, journal writing (27.73 vs 25.21 occurrences) and science (9 vs 4.09 occurrences). The adoption of the apostrophe, therefore, is more reluctant in this type of formal and academic writing than in all the other genres. Although the rendering -'s is, interestingly enough, also widespread in the case of diaries in the period 1650–99 (31.59 vs 14.84 instances), which might have been influenced by the conventions of printing at the time, letters, also representative of handwriting, did not exhibit this until the first half of the eighteenth century (21.32 vs 5.26 instances).

Table 5. The apostrophe in the possessive case across genres (n.f.)

On another note, drama is again of special interest since the apostrophe prevails over the inflected form from the first subperiod (with 33.43 and 20.47 instances, respectively). The early dissemination of the apostrophe in dramatic texts is again connected with its potential to mark the omission of a sound and, therefore, more likely to appear in those texts written to be recited. Similarly, the spread of -'s is also notable in legal texts from the onset of the seventeenth century; these texts were presented to the court for its consideration and their implicit orality stems from the fact that they were in principle written to be read aloud as well. After these subtle differences in the expression of the possessive case in the seventeenth century, the period 1700–49 definitely marks the systematic diffusion of the apostrophe, which thus predominates over the inflected form across all genres.

3.3 The apostrophe in the nominative plural

This is, in our opinion, the most peculiar deployment of the symbol in the history of English writing. In itself, it is not a symbol specifying a grammatical, semantic or pragmatic function, but a practice which at some point plural nouns undertook for visual purposes. The apostrophe in these environments is often referred to as the greengrocer's apostrophe, defined as the placing of an apostrophe before the terminal -s of a nominative plural noun (OED s.v. greengrocer's apostrophe n.), regardless of whether the noun ends with a vowel as in orange's, or a consonant as in doctor's. Some grammarians have claimed that this is the result of lack of education, where writers tend to hypercorrect in those contexts considered to be ‘too abrupt’, as in pie's rather than the correct plural form pies along with numerals such as 1970's, the letters of the alphabet like three A's, and other odd spellings like do's and dont's (Hook Reference Hook1999: 45‒6). Oddly enough, this use of the apostrophe does not share any connection with its original use, elsewhere conceived to mark omitted letters. In this very case, however, there is no <e> to omit in the expression of the plural,Footnote ¹⁴ being instead a redundant symbol reduplicating the function already expressed by the plural morpheme.

The greengrocer's apostrophe dates back to the seventeenth century; it was initially more likely to occur after vowels and later after consonants as well (Cook Reference Cook2004: 94). Table 6 reproduces the occurrence of the plural morpheme both with and without the apostrophe in the corpus where it is shown that this use of the symbol was negligible in the history of English. It is found to reach 1.37 occurrences in the period 1600–49 and since then it has decreased to such an extent that it becomes practically non-existent throughout the nineteenth century, perhaps as a result of the grammarians’ proscriptive pronouncements (Alford Reference Alford1864: 21; Beal Reference Beal2010a: 59). This practice was surely the result of the impulse of some seventeenth-century printing houses and, unlike the previous uses of the apostrophe, it did not diffuse perhaps on the assumption that it did not have a particular clarifying function and could be easily left out, hence the overwhelming preference for the historical morpheme -s.

Table 6. The plural nominative -s and -'s (n.f.)

In view of the low frequency of this symbol from a historical perspective and the grammarians’ proscribing attitudes, one cannot but wonder why it gained such a prominent role throughout the twentieth century. The issue, according to Hook, is not a matter of pronunciation but rather one of representing speech through pronunciation, especially for those writers without an education in English grammar and spelling. On the one hand, it is the result of the interference in those plurals which end in an -s, where the average writer of English does not see the wood for the trees when differenting the boxes and the boxe's or even in family identifications such as the Jones, the Jone's, the Jones’, the Jones's or the Joneses’. At other times, on the other hand, the issue stems from the ‘misstimuli due to frequent examples of errors in the various print media’, which spread the generalisation and socialisation of particular writing conventions like the one at hand (Hook Reference Hook1999: 49). This use of the apostrophe is a rara avis today and the sporadic cases are limited to commercial purposes rather than the true survival of the nominative plural form, which did not require any type of case marking besides the -s termination. As far as the genitive was concerned, however, the apostrophe was required as a marker to differentiate it from plural nominatives.

3.4 Concluding remarks

The apostrophe has been described as a printer's mark especially designed ‘for the eye rather than for the ear’ (Sklar Reference Sklar1976: 175; Little Reference Little1986: 15). Though the symbol was rationalised as marking a phonological elision, its functions in the past and genitive inflections became purely orthographic: the apostrophe simply highlighted the omission of the graph <e> in the morphemes -ed and -es, where the pronunciation of the vowel would only be prompted under particular phonological circumstances. Despite their functional similarities, -’d and -'s had entirely different outcomes in standard English, as discussed above and also illustrated in figure 4. Although the apostrophe was far more frequent in the expression of the past than of the genitive until the first half of the eighteenth century, the former dropped just as the latter was extending across text types and entered standard usage. This turn of events thus raises questions regarding the standardisation of the apostrophe: what led to the rise of -'s and to the decline of -’d? Is there correlation between the two phenomena?

Figure 4. Distribution of the past (-’d) and genitive (-'s) apostrophes over time

The decrease of the apostrophe in the expression of the past tense seems to have been motivated by its own superfluity. The decline of -’d must have run parallel to that of -t and -’t: other than the surplus of forms performing one single function, the frequency of -’d was simply lower and its distributional niche more specific than those of its counterpart -ed, and thus this variant was discarded from usage. This outcome is worth considering against the backdrop of the English writing system, which, despite being sound-based, also exhibits some morphological or morphemic representations. This implies that morphemes with different phonological realisations always have the same spelling, as in the regular past tense -ed or the plural number -s (Cook Reference Cook2004: 18, 75). Conversely, it was the redundant nature of the apostrophe which secured its standardisation in the Saxon genitive. Although the symbol had once stood as an abbreviation for -es, the full form was long forgotten among contemporary speakers, with the apostrophe standing as a vestige of the past (Sklar Reference Sklar1976: 175‒6). However, there was much confusion and a pressing need for disambiguation in nouns ending in -s, since it was often unclear whether the inflection corresponded to the nominative plural or the genitive case, regardless of number. The apostrophe thus became the signpost of the genitive, spelled -'s for the singular and -s’ for the plural.

In themselves, linguistic changes unfolding in the Early and Late Modern English periods often bring to the fore the roles possibly played by either grammarians or printers. The first took upon themselves the fixing of the language and a myriad of grammars were published from the mid eighteenth century onwards. Their timing, however, poses a problem in the present discussion: by the time the grammar-writing tradition became prominent in England, the apostrophe was headed well into extinction in the past tense and into standardisation in the genitive case. The discussion in sections 3.1. and 3.2. illustrated that, though grammarians did still engage with the regularisation of this symbol in the possessive (see section 3.2.), they ignored the construction -’d, perhaps because its decline preceded their endeavours. At any rate, the efforts of these prescriptivists regarding the uses of the apostrophe are best summarised by Leonard (Reference Leonard1962: 197), who claims that ‘[e]ighteenth-century grammarians introduced confusion into usage, and … usage in spite of them settled the matter quietly as it had formerly stood’. The lack of involvement on the part of grammarians, despite overtly commenting on the language, suggests instead a role for the other potential actors of change: printers.

The printers’ influence on the standardisation of orthography was, if anything, indirect. As part of their tasks, correctors, typesetters and printers would tweak the spellings of the manuscript, either to accommodate the text to external needs (Scragg Reference Scragg1974: 72) or to follow specific in-house guidelines (Tyrkkö Reference Tyrkkö2013: 152). Although the mid sixteenth century saw the formation of a printers’ guild (Rutkowska Reference Rutkowska2013: 125‒6), there is no material evidence to corroborate the collaboration between members of the trade: there are no stylesheets or any other usage guidelines pointing towards a joint endeavour (Brengelman Reference Brengelman1980: 333). Nonetheless, recent studies have confirmed that printers did perform as a community of practice in the promotion of certain orthographic forms that eventually became standard (Rutkowska Reference Rutkowska2013: 142; Tyrkkö Reference Tyrkkö2013: 169). It seems plausible that the apostrophe should have been standardised as the product of such communal practices. Interestingly enough, while grammarians were more concerned with the cases of elision of the apostrophe, our data suggest that it was printers who again pioneered and disseminated the use of this symbol for the expression of the singular possessive case in the seventeenth century, which the grammarians would accept in the course of the eighteenth century (Salmon Reference Salmon and Lass1999: 48).

The functions performed by the apostrophe thus coincide with the self-appointed objectives of the early printers. According to Moxon (1638: 220, as cited in Tyrkkö Reference Tyrkkö2013: 157), ‘A good Compositer is ambitious as well to make the meaning of his Author intelligent to the Reader, as to make his Work shew graceful to the Eye, and pleasant in Reading’ and the apostrophe might have been initially introduced in English for the purpose of maintaining these pretensions. In the first place, -’d and -'s were aids for reading out loud, especially fitting in dramatic texts. Though the apostrophe remained relevant in the genitive over time, -’d became superfluous. This shift, along with the appararent desire for regularisation on the part of the printers, might have triggered the extinction of this spelling variant altogether. Although the analysis in this article does not give sufficient grounds to claim that the printers’ community of practice led to its standardisation, the data are open to this interpretation. After all, the apostrophe was diffused in printed texts like drama, law and prose fiction, while handwritten genres like letter-writing jumped on the bandwagon at different points in time. There is, then, correlation inasmuch as the promotion of -'s had the same rationale as the disappearance of -’d.