Modal Verbs in New Zealand English Directives '

The distribution and semantics of the different modal verbs in English have been explored using a number of corpora (e.g., Palmer 1979, Coates 1983, Quirk et al 1985, Biber et al 1999, Kennedy 2002). Will has been found to be the most frequent of the central modal verbs, while will, would, can and could occur a lot more often than the other central modals (Coates 1983: 23, Biber et al 1999: 486, Kennedy 2002).


Introduction
The distribution and semantics of the different modal verbs in English have been explored using a number of corpora (e.g., Palmer 1979, Coates 1983, Quirk et al 1985, Biber et al 1999, Kennedy 2002).Will has been found to be the most frequent of the central modal verbs, while will, would, can and could occur a lot more often than the other central modals (Coates 1983: 23, Biber et al 1999: 486, Kennedy 2002).
To date, research on the modal verbs in New Zealand English has not been conducted, although there has been a widening of our knowledge of this variety of English in recent times (Bauer 1994).An initial search of the one million word Wellington Corpus of Spoken New Zealand English (WSC), shows that would is the most frequent of the central modals.This is followed by will, can and then could.
The prevalence of would in the WSC may be accounted for by the large proportion of informal data in this corpus.The distribution of the different modals has been found to differ between written and spoken texts and according to the exact context.Kennedy (2002), for example, shows how the distribution of the modal verbs varies in the different genres within the written texts in the British National Corpus (BNC).Because there are different types of meanings associated with different modal verbs, the frequency of the modals would also be expected to differ from one type of spoken data to another.Focussing on the transactions and meetings ' I would like to thank the women who so generously recorded the interactions which provided the data for this study.The data is drawn from the Language in the Workplace Project based at the School of Linguistics & Applied Language Studies, Victoria University of Wellington, New Zealand (www.vuw.ac.nz/lals/lwp).I would also like to thank the anonymous reviewer who gave me very useful feedback on an earlier draft of this article. 2In Vine (2001) and Vine (2004) I use the term control act to refer to directives, requests and advice.In this paper, the term directive covers all of these.The broad definition of directive adopted here is "an attempt to get someone to do something".Simple questions for information and clarification are not included.

3
In particular my data involves mainly what I call LATER directives rather than NOW directives, i.e., they are predominantly directives which relate to actions which will be carried out in another place and time.This has implications in terms of the basic form used.I would predict that a corpus of predominantly NOW directives would have a lower usage of modal verbs as there would be a lot more imperatives.section of the WSC, will rather than would is found to be the most frequent of the central modals.
In this paper, I examine the use and meaning of modal verbs in a small corpus of workplace data collected in New Zealand and, more particularly, in situations where women are asking each other to do things, i.e., in directives 2 .I also investigate the use of marginal auxiliaries associated with the different modal meanings.Over the whole of the 52 extract sample analysed, will is the most frequent of the central modal verbs and, as in studies of other varieties of English, will, would, can and could occur a lot more often than the other central modals.This pattern is not repeated, however, when the directive utterances are pulled out and examined.
Coates ' (1983) and Kennedy's (2002) results were observed in large corpora.My corpus of 52 extracts only amounts to 85,268 words and the 439 directive utterances that I have identified in these transcripts total 6,190 words.There are 209 occurrences of centtal modal verbs in the directive data, i.e., 3.4% or 34 tokens per 1000 words, while in the overall data there are 28 tokens per 1000 words.Kennedy (2002: 77) found 21.5 modals per 1000 words in the spoken texts from the BNC.As well as considering the distribution of the central modals, Kennedy (2002) also explored the distribution of need to, ought to, dare and used to.The total of 21.5 modal verbs therefore includes figures for these verbs.Combining my results for need to with the results for the central modals means that I have 252 occurrences across all directive utterances.This raises the number of modals per 1000 words to 41, while the overall figure rises to 30 tokens per 1000 words.Directives 3 , therefore, are one type of speech act which makes a great deal of use of modal verbs, although the workplace context also appears to be one where speakers make frequent use of modals.

The Data
The results presented in this paper come from a larger study exploring the expression of power in interactions between a group of women in a New Zealand government department.Modal verbs were not a major focus of the research but results found in a brief look at modal use as a modifying device in the expression of directives, showed some interesting patterns which did not follow the overall pattern present in the WSC nor that found in other varieties of English as noted by researchers such as Coates (1983) and Kennedy (2002).More significantly, it did not reflect the overall distribution present in the 52 interactions.
The focus of the larger study on workplace language came from an interest in the language that people use at work.Workplace communication has only recently become a focus for linguistic research.Many people spend a large proportion of their lives at work, making this a very important context in which language and communication should be studied.The Wellington Language in the Workplace Project aims to examine real workplace interactions and my dataset.was drawn from recordings made in the first workplace where data was collected.Women's language was analysed because although there were some men in the workplace investigated, it was predominantly a female workplace.The few men there were only recorded on a few occasions.My main focus was on two women managers and the ways their status was or was not evident in their one-to-one interactions with their staff (Vine 2001(Vine , 2004).
As noted above, the data examined comprises 52 interactions and 85,268 words.I identified 439 directive utterances in this data4 .One hundred and five of these were imperatives and 163 were implicit, i.e., the speaker did not explicitly state what they wanted the addressee to do and/or that the addressee was to do it.The results in each section below refer mainly to the use of modals and marginal auxiliaries when these modify the action specifying verb, i.e., in the 171 verb phrases where a speaker explicitly states what they want the addressee to do, although overall results are also given.
I use the terms implicit and explicit in a specific way here.The following utterances are all coded as explicit.There is imbedding at times, but I would argue that the action required and the agent of the intended action are still retrievable from each utterance.
Example 1: a. can you just write that up a bit neater?b. and if you can just later today have a check through that that would be good c.you need to just check the travel booking d. and then urn we could contribute to the discussion that way e. tell her we might be a few seconds Example 1(a) is a conventionally indirect utterance.Although the literal meaning in this type of utterance is about ability, the conventional meaning is directive.In order for the literal meaning to be understood there needs to be something in the context that indicates that this meaning is intended.
The use of the pronoun we in Example 1(d) is interesting because it could be argued that the agent of the action is ambiguous and the directive is therefore implicit rather than explicit.We can mean you, we (the two of us or the two of us and everyone else as well) or I.Because of the workplace context in which my data was collected we often means the organisation -or more specifically in the cases I am talking about we or you as member(s) of that organisation.The job roles and obligations of the individuals involved often mean that the meaning can be more clearly defined.Generally the actions being referred to are the addressee's responsibility and they know this, so the speaker does not need to be more explicit.They do not have to use you as we is explicit enough.In the interaction from which 1 (d) was taken the speaker was giving the hearer advice about an upcoming meeting that the hearer would be attending as a representative of the organisation they both work for.The speaker was not going.
In Example 1(e) the directive takes the form of an imperative 'tell her'.The thing that the speaker wants the hearer to tell a third person is 'we might be a few seconds'.The central modal might is used here, but does not modify the action specifying verb.Modals used in this way are not counted for the explicit directives results below, but are included in the overall figures.
The utterances in Example 2 are all implicit.The focus in this type of utterance is frequently on the speaker, an object or a third person's need or action.In order for these things to be fulfilled however another action is required on the part of the addressee.The meaning in implicit directives may be similar tó the meaning in explicit directives, but these were separated out because of my interest in the relationship between the speaker and the hearer.Reference to oneself, a third person or a thing distances the directive from the hearer and therefore modifies the force of the directive.Some implicit directives can be more forceful than explicit directives, e.g., 2(c) above in comparison to 1(a), but the reference to the object is another strategy which softens the directive in a different way.
The central modal verbs and the marginal auxiliaries in the directive utterances in Example 2 are not modifying a verb specifying the action which the speaker requires of the hearer.The difference between explicit and implicit directives is particularly pertinent when looking at utterances such as Example 2(d).Here there is both a reasonably strong modal, should, as well as a more hedged modal would.It is the softer of the two which is most relevant to the action required, but even then the action is distanced from the hearer.

Results and Discussion
I will now explore the distribution of the modal verbs according to their meanings.The semantic classification used is a fairly simple and traditional one, but is useful in examining the meaning of the modals in the specific directive context involved.Unfortunately there have been no investigations of the meaning of the modal verbs in New Zealand English, although linguists in New Zealand are aware that there may be some interesting differences (e.g.Bauer 1987).
Example 3(c) and (d) are declaratives.In these cases the meaning of will is more one of intention or prediction.The proposed action will take place in another place and time.The speaker's use of will strongly asserts what will happen.
Table 1 gives the number of occurrences of the modal verbs and marginal auxiliaries of volition/prediction.Syntactic construction is not often highlighted in this discussion as my corpus is small.210

Modal Verbs and Marginal Auxiliaries of Volition and Prediction
The most frequent modal found in the WSC is would, while the most frequent modal in the 52 workplace interactions I examined is will.Would and will, along with shall are modals of volition and prediction (Biber et al 1999: 485).Coates (1983: 167) also lists will and shall as modals of volition and prediction, but differentiates between these and would, which is a 'hypothetical' modal (Coates 1983: 205).Biber et al (1999: 485) also list BE going to along with the central modals of volition/prediction.Quirk et al (1985: 229) note that will meaning 'willingness' is a meaning which 'is common in requests and offers'.In requests, this involves the use of an interrogative form.Example 3(a) and (b) are the only two directives in my data which are modal interrogatives containing will. 5   Example 3: a. as soon as you've contacted Yvette will you let me know what the story is? b. will you have time to do that today if I fire the stuff across to you? c. we'll put it with the you know the three separate papers that we've made up d. we'll just say it straight out that most of credits have been in this [topic] sector that you might want to look at in priority so other areas where you would increase or make the access for [social group] There were only 14 occurrences of will, would ana shall modifying the verbs in the explicit directives.Contrary to expectations therefore, would and will were not common, although will was more common than would or shall.
Occurrences of will more than doubled however when looking at the overall results.This is not surprising given will's overall prevalence in spoken data.At times it is used in the implicit directives in a similar way to that found in Examples 3(c) and (d), for instance when the intended state of an object is asserted, this state only being attained after the hearer completes the required action.
Would is an interesting modal because of its use in modifying phrases.Overall there were 44 instances of would in the directive utterances in my data.When the analysis focuses on the occurrences of modals modifying the verb specifying the action in the explicit directives, this figure drops to two.This is because would frequently occurs in the phrase '(that) would be good' as in the examples in Example 4.
Example 4: a. it would be good if you could think about that b. make it sort of later next week would be good The use of would in these phrases reflects the fact that would is perceived as being very polite.James (1978) and Fraser & Nolen (1981) each asked 40 subjects to rank a group of sentences which varied on a number of factors, including the use of different modal verbs.Eight of James' 14 sentences contained either may, would or can.Sentences with may were rated as the most polite, followed by would and then can (James 1978: 180).Fraser & Nolen (1981) explored a wider range of modals and syntactic structures.Once again, would was rated as being weaker than can.It was also rated as more deferent than could (Fraser & Nolen 1981: 101).
Another interesting result in relation to this phrase is that often an if clause was present, of the type found in Example 4(a) above, but the 'that would be good' part was omitted.This was actually more common than utterances where it was included, 15 out of 22 didn't contain the matrix clause.If could be regarded as a directive marker, or even as a politeness marker in the way that please is often interpreted 6 .

Modals of Permission, Possibility and Ability
Can and could arc also frequent in spoken English.Can, could, may, might and be able to can all be used to convey the meanings of permission, possibility and ability (Coates 1983: 85-130;Biber et al 1999: 485).In relation to directives, Coates (1983: 98-99) notes that the use of'CAN= v Root possibility' to imply willingness ... is taken one step further in its use as a covert imperative'.Coates (1983) is referring to the use of can in interrogatives here, but the same point applies to declaratives.In these cases, can questions the preconditions of a directive that specify 'that the addressee should be able (i.e.nothing prevents him carrying out the action) and willing' (Coates 1983: 98).Overall, can and could account for 60 of the 98 occurrences of the central modal verbs that modify the action specifying verb (61%).Coates (1983: 98) notes that in these directive situations, will (and would) 'occur more frequently than' can (and could).In my explicit directive data, can (and could) occur much more frequently.This pattern is evident in both interrogatives and declaratives.Some examples are provided in Example 5.
Example 5: a. can you please make sure that the room is booked for the whole day?
b. could you have a quick word with her? c. you can say that I'm going to send them an example Over all the directive utterances can is also most frequent, although this is then followed by would.Could and will show a similar distribution, although could is slightly more frequent.
Some researchers on directives and requests note that the difference between can and could is one of tense (see for example Trosborg 1994: 210).As Biber et al (1999: 485) note, however, the main function of pairs of modals such as can and could relates to 'speaker stance rather than the marking of time distinctions.For example, modals associated with past time are also associated with hypothetical situations, conveying overtones of tentativeness and politeness'.Coates (1983: 121) also refers to the use of'hypothetical COULD ... as a polite form of CAN.'This is especially true of a directive context.In every case where could is used a future act is required.
Like could, might has a hypothetical meaning when used in directives.Coates (1983: 161)  Might and could had a similar distribution in my data, as seen in Table 2.
Corpus studies have shown that might tends to occur a lot less frequently than could, can, would ana will.This suggests that the directive context is one which favours the use of this modal.This is further supported by looking at the overall distribution of these five modals in the 52 workplace interactions.Overall there are 203 occurrences of might, while the closest of the other four central modals, could, can, would and will, in terms of frequency is can, with 333 tokens.

Modal Verbs and Marginal Auxiliaries Expressing Obligation and Necessity
The last group of modals and marginal auxiliaries are those which express obligation and necessity.These are much less frequent in spoken English than modals of volition/prediction and of permission/possibility/ability (Coates 1983, Biber et al 1999: 493).Of the two central modals which express obligation and necessity, must and should, only should occurs in my data and its use is infrequent (see Table 3).It is interesting that the only central modal of necessity and obligation used is one of'weak obligation' (Coates 1983: 58).Coates (1983: 58-59) notes that 'where the speaker, in subjective examples of Root MUST, demanded action, with subjective SHOULD, he only suggests it.In the case of MUST the speaker expects to be obeyed, but in the case of SHOULD there is no such expectation'.
The utterances in Example 7, were both said by managers.I would argue therefore that the speaker expects to be obeyed.The strategy used, however, is more polite and less forceful because it suggests that the addressee does have a choice.Biber et al (1999: 495) also found that must 'marking personal obligation' was rare in conversation.They concluded that this 'is probably due to the strong directive force this modal has when used in face-to-face interaction.The modal should provides a hedged expression of obligation that is typically regarded as more polite'.Other researchers mention the association of should with suggestion and advice.Altman (1990), for example, explores the interpretation of two forms which he associates with the expression of advice -should and had better.
The infrequent occurrence of should and the non-occurrence of must is interesting given that they are modals of obligation and necessity.It could be expected that in a workplace context where the tasks required in the directives relate to the job obligations of the addressees, that this may be a factor that is referred to through the use of this type of modal.A number of other marginal auxiliaries have also been associated with the conveyance of this type of meaning.Biber et al (1999: 489-490) note that the low occurrence of should and must in their data could be partly explained by the fact that 'semi-modals have become better established in this semantic domain, apparendy replacing the modal verbs to a greater extent'.Need to only accounted for 0.2% of the modal verbs in the BNC (Kennedy 2002).As seen in Table 3, need to is reasonably frequent in my explicit directives.If the results for need to are combined with the results for the central modals, need to accounts for 17% of these modal verbs.This is one form, therefore which the women in my dataset use a great deal when issuing directives.It would appear that when the focus is on directives this marginal auxiliary is common.I do not know if this pattern is peculiar to New Zealand English, or whether it would also be found in other varieties of English.I have yet to explore the distribution of need to in the WSC.
Kennedy is one of the few researchers who has explored the distribution of need to.This reflects the fact that need to has not always been accepted as a modal.Coates (1983: 49) and Palmer (1990: 127) are both careful to distinguish between need the modal and need to the nonmodal.Biber et al (1999: 484), however, call need to a 'marginal auxiliary' and count the marginal auxiliaries (including need to) as 'semi-modals'.
The status of need to as a marginal auxiliary can be justified by looking at the way that it functions.Coates (1983: 31-84) explores the modals of 'obligation and necessity'.She includes the 'quasi-modals' have got to and have to because 'no discussion of MUST or of the modals of Obligation and Necessity would be complete without reference to them' (Coates 1983: 52).Coates (1983: 52-58) shows how these quasi-modals can function in a similar way to must.A similar argument could be applied to need to.The examples in my data with need to also have the 'meaning of MUST' (like have got to and have to), i.e., 'they can be paraphrased 'it is essential that" (Coates 1983: 53).Need to is softer than must, but the meaning is similar.Need to allows the speaker to avoid direct reference to their own authority.Need to implies external forces require the task to be done, and therefore distances the directive from the speaker.
Along with must, should and need to, Biber et al (1999: 485) list a number of other marginal auxiliaries which convey the meaning of obligation/necessity: (had) better, have (got) to, ought to, be supposed to.
Although not frequent in my data (see Table 3) The low frequency of have to is unexpected given the findings of corpus studies.Biber et al (1999), for example, found that have to was used more than any of the other modals or marginal auxiliaries which express obligation and necessity.It may be that in New Zealand English this role has been taken by need to, a marginal auxiliary which is stronger than should, but is weaker than must or have to.

Summary and Conclusion
Directives, in the context in which I have examined them here, are a type of speech act where modal verbs are used a great deal.In the action specifying verb phrases in the explicit directives, the most common modal verbs and marginal auxiliaries used in the directives are modals of possibility.Can, could, might and be able to account for 51% of the modal verbs and marginal auxiliaries.These modals are associated with a low level of force and convey high levels of tentativeness and politeness.
Modal verbs and marginal auxiliaries associated with obligation and necessity account for a further 36% of the modals in the explicit directives.This type of modal can strengthen the force of an utterance, although the strongest modal in this group, must, does not occur at all.Need to was the most frequent of the modals in this group.It is at least twice as frequent as any of the central modals except can.There were ten occurrences of have to.The other marginal auxiliaries do not occur frequently, even though the workplace context is one where reference to job obligations might be expected.
Looking at the overall results, can is still the most frequent of the modals and marginal auxiliaries, with 64 occurrences.Would and need to have similar frequencies, 44 and 43 respectively.Might ranks fourth with 30 occurrences.
These results differ from the overall patterns found in studies of spoken English in a number of ways.In particular, there is a relatively low use of will, while need to has a high frequency in the directive data.The modals that occur frequently generally serve as softening devices, with modals associated with a strong level of force, such as must and should occurring relatively infrequently.The two women managers who uttered the majority of the directives were found to have a supportive style of interaction with their staff.They showed concern for others' face needs and this was evident in the ways they expressed their directives.They were not authoritarian, rather they worked to maintain good relationships through their patterns of interaction.Only 7% of their directive utterances did not have internal softening devices of some type, modals being one such device which was used frequently.
The different patterns of modal verb frequency observed may be a reflection of the variety of English examined.A quick look at the WSC disputes this, however, since the central modals in this corpus do not show the same pattern.The overall data from which the directives have been taken also provide further support to refute this, as the distribution of the modals here also differs.
The focus on women's speech could explain the differences.This may be the case, but without examination of workplace data collected from men it is not possible to confirm nor completely dismiss this proposition.I have not yet examined the directives of any of the male managers that have subsequently recorded interactions for the Language in the Workplace Project.Comparing the overall pattern found in the 52 extract workplace sample I used to another sample of data from the Language in the Workplace which involves men in one-to-one meetings, however, provides evidence which suggests that gender is not an overriding factor.The most frequent central modal in a 181,142 word sample of male interactions was also the most frequent in the overall women's data, i.e., will.The least frequent central modals also rank in the same order as in the women's data -may, must and shall being least common.There is some interesting variation in terms of the ranking of the other modals.Whether this is due to gender differences or other factors, such as the exact type of interactions involved, needs further exploration 7 .
One conclusion that I have reached after the brief investigation presented here is how interesting modal verbs are and how much more research could be done to explore their use in New Zealand English -both in terms of frequency and use -and within a range of types of data.Modality is a challenging area of the English language, which is certainly worthy of further attention Victoria University of Wellington, New Zealand 7 Different types of speech act can be found in different types of interactions.For example, an interaction whose main purpose is task allocation will have a high occurrence of directives, while a report back/information update meeting is likely to have a much lower incidence.Can and need to, for example, may then have a high corresponding frequency in the task allocation interaction, while other modals may occur more in the report back meeting.
I can look at it then okay meaning have it ready for me then b.
notes that the use of hypothetical Root might 'is often used to indicate a course of action politely, without giving overt advice'.Example 6: a. so you might get them to score their own work as to the extent to which it satisfies the criteria b. you might like to just reassure them c. you might wanna rewrite it should probably put in there that um the ministry has what we did actually intend need to add in a column or something b. and once you've faxed it through we need to send them th-the original c. you need to just check the travel booking , there were 14 examples found in the explicit directives.Some examples are provided in Example 9. Example 9: a. you're supposed to be making an appointment b. we have to fax this off c. and we have to arrange fifty percent payment d. but we better take that one with us now I need to get that up to them today meaning get back to me quickly on this so I can send it off c. that needs to be couriered up to [name] today meaning get that couriered to [name] today d. if you've got views on how you think it should be structured um that those ideas would be really helpful meaning give me feedback on this e. he could be a useful ally meaning get in touch with him

Table 1 :
Occurrences of modals of volition/prediction

Table 2 :
Occurrences of modals of permission, possibility and ability 6Chris Lane (personal communication)

Table 3 :
Modals and marginal auxiliaries of obligation/necessity