Historical Sociolinguistics and Sociohistorical Linguistics

Home ] Up ]


What speakers’ gender, age, and native language
reveal about their notions of English usage norms

Thomas de France (contact)

(University of Leiden)

Received: May 2008, published April 2010 (HSL/SHL 10)

1. Introduction

1.1 The influence of early English prescriptivism
Discussions, debates, and revisions over English usage today owe much to the work of early codifiers, in particular those who attempted to lay down not only the specific rules of usage but the elaborate justifications for such rules. Since its inception in the eighteenth century (Mittins et al. 1970; Peters 2006), the English prescriptivist movement has certainly maintained its share of fans up to the present day. Anyone who may doubt whether the influence of the grammarians of old still holds sway need only look at a sampling of the general public’s attitudes toward certain constructions, not to mention vocabulary and pronunciation. Time and again it has been shown that many constructions considered to be flat-out wrong today were in common usage until suddenly incurring the wrath of some authoritative, prescriptivist writer.

1.2 What a survey can tell us about usage norms
Usage surveys have proved an insightful tool in elucidating what individuals of various social strata deem acceptable and taboo, both in the spoken and written varieties. By means of a survey, Mittins et al. (1970) gathered data relating to 457 English speakers’ perceptions of various usage norms. The study’s purpose was to ascertain “a notion of the general character and distribution of views on acceptability of usage” and to put “current attitudes and judgments in a historical context ...” (1970:3). The participants responded with their impressions of 55 popularly debated items, either accepting or rejecting each item in a total of four contexts: informal speech, informal writing, formal speech and formal writing.

A different sort of survey, but one done in much the same spirit, was conducted by BBC Radio 4 in 1986. At the request of the program English Now, listeners submitted their complaints regarding current usage, much of it apparently perpetuated through the BBC airwaves. Crystal (1995) published the top-ten errors sent in by those who responded to the radio program with their complaints. Tieken-Boon van Ostade (2005) makes use of the list, putting the items into questionnaire format and administering the form to participants present at the annual Federation of Finnish-British Societies. Finn-Brits members ranked the ten items on the survey in order of acceptability, assigning one rank per item. They had been asked not to spend excessive time thinking about their choices, but simply to give their first reactions. 

1.3 The current study
The University of Leiden was the site of a more recent usage survey. The date was 3 November 2007, and the occasion an alumni meeting open to all graduates of the Department of English Language and Culture. A total of 80 survey forms were filled out and returned. This article examines the results of the Leiden survey in light of those of the Finn-Brits study.

The form used was the same as in the Finn-Brits survey, though this time informants were also asked to fill in background information, such as age and native versus non-native speakership. The purpose of the present study, therefore, is to serve as a continuation of the Finn-Brits survey, while also attempting to find correlations between respondents’ notions about English norms and sociolinguistic data that make up the speakers’ backgrounds.

2. Literature

2.1. History
First it is necessary to present a brief history of the ten usage items that made up the Leiden survey, including when the usage was first thought to appear, when it subsequently fell into disfavour and why.

2.2. Ten debated items
Double negation. Many an uninformed English grammar stickler may be surprised to learn that for centuries “double negation was the rule in all its varieties ...” (Tieken-Boon van Ostade 1995:12). Burchfield (1996) suggests that its origins can be traced back to Old English, in which an additional negative particle simply added more emphasis. When and why its eventual stigmatization, then? It was, according to Burchfield, at some point between the sixteenth and eighteenth centuries, although for reasons unknown, that double negation became “socially unacceptable” (1996: 227). The proscriptive attitude may well have been the outcome of the written word as an ever-increasing medium of communication in the fifteenth century (Tieken-Boon van Ostade 2005), owing to the recent invention of the printing press. In the latter part of that century, Tieken-Boon van Ostade (2005) elaborates, the practice of using two negative particles in the same sentence perhaps began to be viewed simply as redundant, as cancelling each other out. Moreover, unlike with spoken language, which can often benefit from additional particles for the purpose of emphasis, the written language tends in direction away from repetition (Tieken-Boon van Ostade 2005).

None with a plural verb. The logic of verb concord is often the justification given as to why a verb should be in the singular when it follows none. Tieken-Boon van Ostade (2005) searched the British National Corpus (which includes spoken usage as well as written) and http://Google.com to find out how frequently the plural form of both the past and present tenses of the verb to be appeared with none. In the British National Corpus, incidents of none is/was were nearly double that of none are/were. Even so, Burchfield (1996) states that there is not necessarily any reason to assume that the singular verb is grammatically less correct. He explains that although the common assumption may be that none means “no one”, the word can actually be traced back to the Old English nān, meaning “none” or “not one”, and that a plural verb was used with nān as far back as the ninth century (1996:526). 

Split infinitives. Although the split infinitive can be traced back to the fourteenth century, so far as we currently know it was not until the nineteenth century that prescriptivists first took a prohibitive stance against the construction (Beal 2004:111-112; Burchfield 1996:736; Tieken-Boon van Ostade 2005:11). Nevertheless, split-infinitive defenders down through history have included such distinguished writers as theologian John Wycliffe (Mittins et al. 1970:70), poet Samuel Taylor Coleridge (Mittins et al. 1970:70), and author/playwright George Bernard Shaw (Crystal 1995:195).

Who for whom. The confusion of when to use the subject and object forms of who dates at least as far back as the fifteenth century (Burchfield 1996:847). Who is, in fact, the only relative pronoun in Modern English that distinguishes case (Burchfield 1996:847). The direct and indirect object forms are both whom, but this one instance of a declinable relative pronoun, in a language where case declensions are almost non-existent, seems to bemuse its speakers. This can be observed in situations when, through hypercorrection, speakers often use whom in place of who. Crystal (1995:194) adds that nowadays whom carries a sense of formality.

Different to/than. The forms different to and different than have often been regarded as inferior to the form different from. All three variants have a respectably long history: the uses of different from and to date back to the sixteenth century, and that of different than to the seventeenth (Burchfield 1996:212). Different from is the most frequently occurring of the three variants (Crystal 1995:451). Burchfield also points out that, at least in British English, “a marked preference for different from has been shown ...” (1996:212). The reason for the preference may be an attempt to remain faithful to the origin of different, in which the Latin prefix dis- means “from” (Crystal 1995:194).  

Dealing first with different to, we find that diverging views exist as to its distribution. Mittins et al. (1970:83) list several authorities who have claimed that different to occurs primarily in British rather than American English, while Burchfield states: “In both countries, in all kinds of circumstances, different to has been widely perceived as a credible alternative ...” (1996:212). Perhaps the different to usage occurs, as Crystal posits (1995:194), due to its similarity to such phrases as “similar to” and “opposed to”. Those who question its legitimacy, though, often do so on morphological grounds; when using the verb to differ, the preposition that follows is from, and therefore, it is argued, this rule should also apply to the adjectival form and yield different from (Burchfield 1996:213). Yet Burchfield (1996:213) also notes that incongruence of this nature occurs often in English: according to but accords with, for example. Others opposed to the use of differ to claim that to “contradicts the etymological force of the dis- prefix constituting the first syllable of different ...” (Mittins et al. 1970:82).

The distribution of different than is equally contested. Burchfield states that this construction, though quite common in American English, “does not form part of the regular language in Britain” (1996:213). Crystal, however, upon examining the Longman/Lancaster Corpus, which includes published texts from English-speaking countries throughout the world, reported that one-third of the occurrences of different than come from British sources (1995:451).

Preposition stranding. Grammatical logic may not be of much help in answering this question, for it seems that out of nowhere the construction suddenly became taboo in the seventeenth century. And apparently all this is thanks to the influence of one man, prescriptivist-avant-la-lettre John Dryden (1631-1700) (Crystal 1995:194; see also Yañez-Bouza 2006). Prescriptivists have often tried to pattern English usage after Latin grammar (Crystal 1995; Peters 2006), as appears to be the case with Dryden and his admonition against the stranded preposition. Crystal defends preposition stranding on the grounds that  “… the preposition, which the verb usually takes after it, is regarded as forming a part of the word itself. To speak of, to resort to, are hardly verbs and prepositions, but form in each case almost one word” (1995:367).

Will with the first-person subject to express future time. The use of will in the first person, as was often vehemently argued by grammarians of old, should be reserved to express “determination or insistence” (Burchfield 1996:707), whereas shall is to be used to indicate future time. The rules are what Burchfield aptly describes as “immensely complicated” (1996:707), because they imply an inversion of sorts, whereby shall is to be used in the second and third person for the same reasons that will applies to the first person. This rather counter-intuitive practice probably started in about the seventeenth century, as Mittins et al. explain, but “the most elaborate formulations flourished” one century later (1970:98; see also Tieken-Boon van Ostade 1985). Mittins et al. call the will/shall set of rules a “monstrous prescription” (1970:98). Crystal (1995) and Peters (2006) doubt in fact whether these complicated rules were ever consistency adhered to at any one point in history. Thus the construction may well have been as good as doomed from the beginning. It appears that the first-person use of shall is fading quickly from British dialects and it is almost absent entirely in United States, Ireland, and Scotland (Crystal 1995).

Hopefully. There are two separate points to be considered with this item: the use of hopefully as a so-called “sentence adverb” and the use of all other -ly adverbs used in the same way. Burchfield explains: “... in the 20c. there has been a swift and immoderate increase in the currency of -ly adverbs used to qualify a predication or assertion as a whole. [...] Suddenly, round about the end of the 1960s, and with unprecedented venom, a dunce’s cap was placed on the head of anyone who used just one of them — hopefully — as a sentence adverb” (1996:702). Crystal, who prefers the term “disjunct” to define -ly adverbs used in such a way, admits that “[i]t is unclear why hopefully has been singled out for criticism” (1995:194). Indeed, the use of -ly adverbs in this way has been common practice since the seventeenth century (Burchfield 1996:703), so the general phenomenon, regardless of the specific -ly adverb being used, has a respectable history.  

Placement of only. The correct position for only, it has been argued, is directly next to the word it modifies. In speech, the context of the utterance, not to mention its emphasis and intonation, generally makes clear what only is intended to modify (Burchfield 1996). The area where greater care is needed, though, is in writing, where such prosodic information cannot be conveyed (Crystal 1995:194). Robert Lowth (1710-1787) was first to comment on the (mis)placement of only, and thus it is since the eighteenth century that the discussion has been ongoing (Burchfield 1996:551).

Between you and I. This usage would seem to be the result of hypercorrection. Crystal points to a lack of awareness as well as a misunderstanding of traditional grammars: “Many educated people are unconsciously aware of the way these grammars have criticized me in other constructions, recommending It is I instead of It is me. They have a vague feeling that I is somehow the more polite form, and thus begin to use it in places where it would not normally go” (1995:194). Burchfield comments that “the only admissible construction … in standard use in the 20c. is between you and me” (1996:106).

What most of the items above have in common is that although they have all received their fair share of criticism from grammarians, they were in general use and presumably used by males and females of differing ages and classes long before their uses became taboo. In many cases, the only reason a given usage is considered a taboo construction today is that one individual deemed it inappropriate and eloquently dissuaded speakers from its use.

3. Method of analysis

Before the usage data from the surveys could be analyzed, it was necessary to determine which of the forms were eligible to enter into the final tally and which were not. As is the case virtually anytime surveys are administered, not all participants follow the directions and/or provide complete information. Following are the issues that arose regarding the filling out of forms.

Of the 80 forms turned in, a total of fifteen participants did not follow the directions of ranking each item only once. Instead, they assigned the numerical values 1 through 10, often giving the same value to several items. Thus these surveys had to be left out of consideration.

Other forms were disqualified for the following reasons: there were cases in which participants’ low ranking and high ranking items contrasted with their peers’ average rankings to such a degree that it seemed they must have filled out their surveys backwards, assigning 1 for least acceptable and 10 for most acceptable usage. These forms were judiciously set aside.

One of the items, the placement of only, proved confusing for a number of participants, most likely owing to the sample sentence: “I only say Jane”. Crystal’s (1995) sample sentence, however, is “I only saw Jane”. It could be that the much smaller likelihood of the occurrence of the former sample sentence, using the verb say, caused respondents to be confused. Many of these respondents, however, had filled out their surveys incorrectly in the first place, and so they had already been excluded from the count.

Three forms presented what seemed to be a “ranking by committee” of sorts. They were suspect because they were apparently handed in consecutively (assuming the order of the forms had not been shuffled from the time they were returned) and all showed an identical ranking order. It appeared that the three participants had considered together all the rankings and had come to an agreement on the survey as a whole. But because the instructions were for each individual to give his or her first reaction to the items, the three forms were deemed invalid.

In total, 57 surveys were considered eligible, although some participants left no response to the questions of age and gender. The available pool in examining gender and the category of native speakership totaled 41. Two age groups were compared, one group of individuals in their twenties and another in their sixties. To analyze the results, the raw data were entered into a spreadsheet so as to calculate the sums of the rankings of each of the ten items and find the rankings of the groups.

4. Results

4.1 Ranking of usage items by all participants
First presented are the results obtained from all 57 eligible surveys. Table 1 shows a comparison of the Leiden results with those of the Finn-Brits participants from Tieken-Boon van Ostade (2005). The higher the ranking position, the less acceptable each item was deemed in comparison with the rest of the items presented. The number of Finn-Brits participants is not known. Therefore, the numbers in parentheses are not to be compared across the two pools of speakers, but are meant as a reference for comparing items within each speaker pool.

Table 1 Comparison of Leiden and Finn-Brits acceptability rankings.





Will + 1st person (186)

Will + 1st person (173)





Preposition stranding (210)

Preposition stranding (202)





Hopefully (222)

Hopefully (203)





Placement of only (267)

None + plural (229)





None + plural (268)

Placement of only (313)





Split infinitive (310)

Split infinitive (319)





Who for whom  (314)

Who for whom (354)





Different to/than (345)

Different to/than (382)





Between you and I (429)

Between you and I (421)





Double negation (531)

Double negation (539)

The most notable feature overall is that the rankings from the two pools of respondents are virtually identical. At one extreme, both pools of participants were least bothered by the first-person use of will to express the future tense; the double negative, on the other hand, was deemed least acceptable. The only disagreement of rankings between the two pools was with the placement of the adverb only and the use of none with a plural verb, although the difference is only by one point among the Finn-Brits members.

4.2 Gender
Achieving an even distribution between the sexes was not possible with the pool of Leiden participants. Of the 57 surveys, sixteen of them did not specify gender. Although that figure still allows for a reasonable number of females (33), we are left with a mere eight male respondents. Table 2 shows the results of the gender variable.

Table 2  The role of gender.



Female Group (N=33)

  Male Group (N=8)



Preposition stranding (109)

  Will + 1st person (20)







Will + 1st person (116)





  Preposition stranding | None + plural (28)



Hopefully (127)








None + plural (133)

  Hopefully (32)







Placement of only (170)

  Placement of only (38)







Split infinitive (186)

  Split infinitive (41)







Different to/than (210)

  Different to/than (57)







Who for whom (216)





  Who for whom | Between you and I (60)



Between you and I (244)








Double negation (305)

  Double negation (76)


Female speakers found preposition stranding to be more acceptable than I/we will, unlike the case with the male group or the speakers as a whole. The other minor difference was a switched ranking between hopefully and none with a plural verb. Males found the latter to be more acceptable, whereas females preferred the use of hopefully.

4.3 Age
Next we are interested in finding out what effect age has on an individual’s perception of English norms. The range in age of the 57 participants was 23 to 74 years, with a median of 44 years of age. This range made the distribution relatively even. In Table 3, two groups are compared: one group in their twenties and another in their sixties.

Table 3. The role of age.



  20s Group (N=12)

  60s Group (N=11)



  Hopefully (33)

  None + plural (35)







  Will + 1st person(37)





  Will + 1st person | Preposition stranding (36)



  None + plural (46)







  Preposition stranding (51)

  Hopefully (45)







  Split infinitive (63)

  Placement of only (49)








  Split infinitive (51)



  Different to/than| Who for whom (77)





  Who for whom (78)







  Placement of only (81)

  Different to/than (83)







  Between you and I (84)





  Between you and I | Double negation (96)



  Double negation (111)



Several oddities are quickly evident between the two generations of speakers. First, whereas the 57 speakers as a whole (not to mention the Finn-Brits respondents) found I/we will to be the most acceptable, neither the twenties nor the sixties age group agreed. The younger members found hopefully the most acceptable, while those in their sixties felt least bothered by the use of none with a plural verb. In both cases these scores are close, yet the results of the Leiden survey overall and the Finn-Brits survey shows the second rank trailing far behind will. Furthermore, not only does each of the age groups differ from the average, but each group also differs from the other. Although the twenties age group ranked hopefully first, the sixties age groups put this item at fourth place. Likewise, the younger groups placed none with the plural verb two positions less acceptable than did the older group.

The placement of only is another interesting case. The twenties age group found it far less acceptable, ranking it eighth. Again, this may be due to the wording of the sample sentence, “I only say Jane”. Yet it is unclear as to why only a younger group of respondents would be particularly sensitive to this. Finally, the double negative was clearly the choice considered least acceptable to the younger group but was tied among the older generation with between you and I.

4.4 Nativeness
It is perhaps not often that an English usage survey is administered to non-native speakers. Although this had also been the case with the Finn-Brits survey, participants had not been asked to supply that information. The Leiden survey provided this opportunity. Unfortunately, not many native speakers present, so the data relating to their attitudes is not as ideal for a control as one would hope. Table 4.4 compares the two groups.

Table 4.  The role of native speakership.



Native English speakers (N=5)

Non-native speakers (N=36)



Will + 1st person (12)

Will + 1st person (124)







Preposition stranding (14)

Preposition stranding (123)







None + plural (20)

Hopefully (134)







Placement of only (24)

None + plural (141)







Hopefully (25)

Placement of only (184)







Different to/than (30)

Split infinitive (187)







Between you and I (28)

Different to/than (237)







Who for whom (32)

Who for whom (243)







Split infinitive (40)

Between you and I (276)







Double negation (50)

Double negation(331)


It seems that native speakers are less patient with the usage of hopefully in place of alternatives such as “I hope that”. They are likewise much less tolerant than the non-natives when it comes to the use of the split infinitive. We can also see that in native speakers hopefully is nearly tied with only, which would move it one position closer to where the non-natives placed it. The figure regarding the split infinitive is more reliable, as it is eight points removed from the higher ranking who for whom construction. As was the case with the gender variable, the two groups of speakers are, unfortunately, lopsided as the native group contains only five people. It is best to consider the above results very cautiously.

5. Discussion and conclusion

5.1. Overall impressions
This study has set out to answer the question of what role three sociolinguistic variables play in speakers’ attitudes toward the acceptability of commonly debated English usage constructions. The three variables, gender, age and nativeness, do indeed reveal certain divergences from the norm.

Of the ten items in the questionnaire, the data show that the most acceptable construction can differ depending on the speaker’s gender and age, though not necessarily depending on whether the individual is a native speaker. Female speakers in the Leiden survey showed more tolerance for the stranded preposition than the typical first-ranking item, I/we will. The age variable is seen diverging from the norm among both of the age groups examined: those in their twenties and sixties. The younger generation shows a preference for hopefully, while the older generation is most accepting of none with a plural verb. Natives were seen as less tolerant than the average speaker with respect to two items, hopefully and the split infinitive. In almost every case there was a clear proclivity to consider the use of the double negative as the most stigmatized. The only exception to this was the sixties age group, who put the item on equal standing with between you and I.

5.2 Ideas for further research

Although the intent of this study was to elaborate on the Finn-Brits survey by taking into account speakers’ gender, age and nativeness, the pools of male speakers and native speakers were, unfortunately, smaller than desired if we want to consider the data as representative. A future study with more respondents would be able to confirm whether the findings in this study still hold true.

Slight changes could also be made to the methodology. In order to gain insight into the differences in written versus spoken norms, respondents could be requested to rate items on these two separate bases. Finally, while instructions asked speakers to assign one ranking position to each of the ten items, this method has its disadvantages. First, just because, for instance, speakers found I/we will to be the most acceptable of the listed items does not tell us how acceptable in general this item was found to be. Perhaps a given speaker has a general dislike of all of the items, and another is not bothered by any of them, not even the double negative. A solution would be that participants would rate each item on a scale of 1 to 10 based on its own terms. Each item could be then judged from an absolute starting point rather than in relation to the other items.


Beal, Joan C. 2004. English in modern times. London: Arnold.

Burchfield, R.W. 1996. Fowler’s modern English usage (rev. 3rd ed., 2004; first ed.: H.W. Fowler, 1926). Oxford: Oxford University Press.

Crystal, David. 1995. The Cambridge encyclopedia of the English language (repr. 2003). Cambridge: Cambridge University Press.

Mittins, W. H., Mary Salu, Mary Edminson and Sheila Coyne. 1970. Attitudes to English usage: An enquiry by the University of Newcastle upon Tyne Institute of Education English Research Group. London: Oxford University Press.

Peters, P. 2006. English Usage: Prescription and Description. In B. Aarts and A. McMahon (eds.), The handbook of English linguistics. Malden, MA: Blackwell Publishing Ltd. 759–780

Tieken-Boon van Ostade, Ingrid. 1985. ‘I will be drowned and no man shall save me’: The conventional rules for shall and will in eighteenth-century English grammars. English Studies 66. 123–142.

Tieken-Boon van Ostade, Ingrid. 2005. “Bad English and the grammarians”. Finn-Brits Magazine Autumn. 8–14.

Yáñez-Bouza, Nuria. 2006. Preposition stranding and prescriptivism in eighteenth-century prose. Historical Sociolinguistics and Sociohistorical Linguistics (go to Contents, Articles).