Printer Friendly

The phraseology of public international English.


This paper has developed from Howarth 2002, which made a preliminary investigation of the language of NATO press conferences. In particular it attempted to discover relationships between discourse features of such a genre (e.g. the way in which journalists ask questions to probe for information) and linguistic features at a phraseological level (such as repetition between speakers and lengthy recurrent sequences).

The earlier paper made use of a corpus of press conferences, which had, however, no mark-up for discourse features that could be used autoniatically by concordancing software, which might allow, for example, independent searching for the speech of journalists or spokespeople. There was, therefore, a limit to how much a standard computerized corpus approach could be followed. There was also great concentration on describing the phraseology for its own sake rather than using it for the understanding of a particular genre of public speech.

The aim in this paper is to deepen understanding of this public genre, again through an analysis of its phraseology. Since the earlier study was conducted, a significant contribution to this field has been published: The Linguistics of Political Argument (Partington 2003). The author of this monograph uses, among other data, a corpus of press conferences held at the White House in Washington, conducted by the press secretaries appointed by President Clinton. Partington adopts as far as possible a corpus approach and is similarly interested in identifying significant discourse features of the genre.

The niain lesson from Partington's methodology (some of the details of which will be discussed below) is the pragmatic solution to the problem of marking up a corpus for discourse studies. In the absence of detailed tagging, it is possible (and much quicker) to split the material into sub-corpora (e.g. for spokespeople and journalists), which can be analysed separately and compared for the presence of certain features of possible significance in the corpus as a whole.

In the following study, I will describe the key features of the NATO corpus and discuss how it has been produced and how much can be known about its creation. I will then discuss how best to analyse the data to extract features significant to this type of discourse. The results of this analysis will finally be put into the wider context of studies of political discourse. The focus throughout will be on answering the question: "How does the conventional language of a social group contribute to the conduct of public, international discourse?"


The data for this study is taken from a corpus of 836 separate press conferences, conducted between May 1996 and September 2002, totalling 2.6 million words. These conferences have been held in Former Yugoslavia, or for a while in Brussels, by the international peacekeeping organizations working in Boznia-Herzegovina and later in Kosovo, operating under the administration of NATO. The material was originally collected for the purposes of professional training for non-native English language instructors in military academies in Eastern and Central Europe. This variety of English was considered a useful example of authentic international peacekeeping English, which, under normal circumstances, is hard to acquire. The original aim was to help them make use of authentic speech for materials design and become more aware of the conventions of this form of communication.

These conferences began in the early 1990s and continue to be held to the present. At times the conferences have been intermittent (every few days or less frequently), at times of great political activity or crisis twice a day. Since 2002 they have been much rarer: barely one a month, so the period of collection of this Corpus covers the peak of production. They are transcribed immediately, and the text is then transmitted via email from NATO headquarters to whoever subscribes to the relevant distribution list. They are also archived on the NATO website as bilingual English/French texts. They can therefore be treated as in the public domain, an added bonus for the researcher.

It is useful at this stage to have an idea of the style of language and the overall structure of these conferences:

<TRANSCRIPT: Joint Press Conference 24 November 1998, 11.40 Hours Coalition Press Information Center Tito Barracks>

<Chris Riley, OHR:> Good morning, ladies and gentlemen. Welcome to the press conference. It's nice to see some old, friendly faces out there--and some friends from Mostar, here--so, I feel quite at home. [...] That's all I've got, this morning. Ariane?

<Ariane Quentier, UNHCR:> I would like to introduce Nicholas Morris. Many of you know him, already--he is our Special Envoy for UNHCR in the region--[...]

<Nicholas Morris, UNHCR Special Envoy:> Good morning, everyone. Thank you, Ariane. I just wanted to give you a short account of this meeting and will make available copies of the High Commissioner's opening statement--and her summing up--during the meeting [...]

<Chris Riley, OHR:> Thank you, Nicholas. Nicole?

<Nicole Szulc, OSCE:> Good morning, everybody. A seminar on military support to civilian authorities will be held in Banja Luka from the 29th of November to the 1st of December. [...]

<Chris Riley, OHR:> Thank you, Nicole.

<LCdr. Glenn Chamberlain, SFOR:> Good morning, from SFOR. Since Nicole and the OSCE feel free to make military announcements--as she did with her first one--SFGR is goinq to make a business announcement. [...] That's all from SFOR. Thank you very much.

<Ariane Quentier, UNHCR:> I will just add this one announcement. UNHCR will support the organization of a concert with a band from Belgrade, Party Breakers, a band from Banja Luka, Revolt, and a band from Sarajevo, Pi Pi Duga Carapa. [...]

<Chris Riley, OHR:> Questions?


<Nedim Dervisbegovic, Reuters, Q:> Do we have any signs of Albanian refugees, in Bosnia, going back to their homes? I believe many of them have expressed their hope that they will be able to go to their homes soon. Have any conditions been created for them to go home?

<Nicholas Morris, UNHCR Special Envoy:> We are only aware of a very few individuals who have requested to repatriate to Kosovo Province. UNHCR's policy is that we are not promoting repatriation--we are not encouraging people, urging them to go back--for very obvious reasons. [...] We owe that to the people who have been waiting and who are still determined to go back home. We also owe it to those who have already made their minds up that maybe they will not go home.

<Chris Riley, OHR:> No more? Thank you very much.

Each conference has an average length of about 3000 words, though they vary considerably. There are those which appear to be purely routine, taking place because they are scheduled (the shortest is less than 200 words). These consist of brief reports from the podium ("We've got nothing significant to report") and sometimes no follow up from the floor: "Questions? No questions? Not a single question? O.K., thanks." On the other hand there are much longer sessions (8,000 words or more), occurring especially when visiting politicians, such as Javier Solana, NATO Secretary-General, are addressing the press corp. On these occasions, we see longer set-piece speeches from the podium and more extensive questioning from senior journalists.

It is not clear exactly what the transcription process involves. Over the years the conventions have changed and there may be more than one transcriber working concurrently. This introduces inconsistency, which may not matter for the principal users of the materials (presumably media organizations), but means that autornatic processing and analysis is made difficult. This may include small alterations in punctuation or larger differences in the conventions for the narning of speakers. The following are all used to indicate a questioner from the floor:










Jamila Milovic, Radio Mir, Q,

(Mike O'Connor--New York Times) Q:

Mrs Savic:

Robert Dania, Kalamazoo Gazette, Q:



audience member:

While a Corpus linguist is fortunate to have this quantity of free, transcribed data, it is frustrating that no standard stylesheet was implemented. Errors in transcription have also been introduced, some of which suggest the work of a non-native transcriber:
who was baged by SFOR       [bagged]
if your interested          [you're]
local compacity             [capacity?]
let's bare in mind          [bear]
after pausing his questions [posing]

In addition to these (assumed) transcriber errors there are examples of the non-native English of some of the conference participants:
if you like to phone me later       [you'd]
and I lost my track of              [begin strike through] my
                                     [end strike through]
you are not in quarrel with ...     [have no quarrel with]
under the authorities of            [authority]

These are of more interest for the present study, can mostly be distinguished from the transcription errors and have been left uncorrected. Sometimes [sic] is used when the transcriber feels the need to make the distinction clear: "why don't you rose <[sic]> it some months ago" (my angled brackets added to the transcriber's square brackets).

These features of the transcripts mean, firstly, that a lot of manual processing had to be done, going through ,each conference to mark off the speakers' names between <>, so that they can be ignored by the concordancer. Secondly, it has to be accepted that an unknown quantity of typographical or lexical errors remain after those found by chance have been corrected. This would clearly undermine a quantitative approach. For example, of 57 instances of BEAR in mind, three are realized as bare. On the other hand, transcribers have (on more than 80 occasions) included 'laughter' or 'scattered laughter', 'laughter in the audience' and 'laughter still'. This may have been done on an ad hoc basis and cannot be taken as a reliable indication of absolute quantities of mirth, but it gives a very useful insight into the kind of interaction taking place.

There is an overlap in genre and subject matter between this corpus and Partington's White House material. Both his corpus and the present one involve participants in high-level contemporary political issues (international and, in the case of the White House, domestic) and, more specifically, cover the years when events in Kosovo were a major international concern. However, there are important differences between the two. The White House data seems to be much more homogeneous, with a more reduced cast of characters: the current press secretary as master of ceremonies addressing the press corps at a daily event. The NATO conferences during the period studied here have a large and revolving set of spokespeople in the chair, with several speakers on the podium and an unknown total number of journalists. The international organizations include the full range of those active in the region post-1990: UNHCR, the EU, SFOR, KFOR, OSCE and others. Furthermore, these NATO spokespeople are mostly non-professionals in the public relations field (chiefly ranking military men and women who have taken on a PR role), untrained in the strategies of information management. There are a couple of interesting exceptions to this rule. Jamie Shea, originally a British academic, became a prominent figure as NATO spokesman and demonstrated considerable skill as the public face of the organization during the events in Kosovo. Mark Laity represented the BBC in many earlier conferences, but, by some time before 2001, had crossed the floor and become Personal Adviser to the Secretary General of NATO and Deputy Spokesman. Many of them are non-native speakers of English, though it is not always clear whether someone called Kris (sometimes Chris) Janowski, for example, is a near-native foreign speaker or native. Finally, the press corps includes several who are clearly non-native speakers and in some cases lack the polished nuance of expert information gatherers.


The contribution that Partington's study makes to an understanding of the mechanics of press conferences is chiefly in identifying the linguistic features that expose the relationships among the discourse participants. In order to do this he recognizes that a purely quantitative analysis of his corpus will not suit: it is necessary for the linguist to be able to move between the different 'levels' of analysis: from the raw audio recordings transmitted over the internet, to the whole corpus (of 250,000 words) and down to smaller sub-corpora. This is in order to "reintroduce as much attention to context as possible- including both the wider historical, political, mediatic context and the more immediate local and personal contexts, especially the participants involved, the speakers and hearers" (4). In this approach he follows such studies as Stein and Quirk (1991), who demonstrate the need to read (and in this case listen to) a corpus, not just rely on what the concordancer is able to extract automatically, which must inevitably be limited by the search strings inputted. Partington claims that this more manual method is more applicable to "specialized corpora, corpora of a single text-type, where these processes and the contexts they take place in remain relatively constant, or at least alter in relatively predictable ways" (ibid). This principle was also followed in Howarth 1996 and 2002 and will be further expanded below.

In the type of discourse he is examining, questions and responses are the 'bed-rock'. However, quite how the Q and R moves are distributed is of great interest. The normal relationship in institutional talk is between the more powerful, professional, questioner and the less powerful, lay, responder (as with doctor-patient dialogue or in the classroom). In the press conference both sides are professionals and it is not certain, especially in the NATO data, that the spokespeople are in fact the more powerful. There is a complex set of factors including who holds the information (the 'gatekeepers'), who has authority to release it to the public and who is the host of the event (possibly requiring a degree of politeness from the guests). In the NATO briefings there are occasions when the journalists know more than the officials, there are externa1 authorities that have to be considered and there are examples of a democratic approach to the timings, for example, with journalists being asked to vote when they would like the next briefing to be held. Partington regards the White House press briefings as informal in comparison with the courtroom or TV interview, and the kind of personal language found in the NATO data confirms this. The spokespeople speak for themselves ('I') as well as for the organization they officially represent, and the kind of phraseology they use reinforces this (e.g. "Where I come from ... ")

Stenstrom (1994) presents a useful set of criteria for understanding this kind of dialogue, for example, how much of what takes place and who speaks is predetermined being significant in institutional talk. A further criterion that distinguishes between different types of talk is whether the questioner already knows the answer. As with courtroom cross-examination or political interviews, the journalist who asks a question in a press briefing to which he knows the answer may well be trying to coerce spokespeople into admitting something they would rather keep under wraps. I'artington calls this function of questioning 'argument organization', which is used for the benefit of the questioner him or herself and the audience. He also makes the important distinction between a response to a question and an answer. A response may not be an answer that will satis@ the questioner; it may be an evasion, of which there are numerous examples in both corpora. The kind of concordance searches that will identify these features expressed in conventional language will be discussed below.

Questions of politeness (in Levinson's (1983) sense) are encountered throughout. The speaker at the podium may make evasion explicit in order to save face: "I'm not going to be led into any discussion about what begot what" (NATO). Positive politeness is at times a useful strategy ("Good question. I have failed you"). Humour is used to disarm the interrogator, as is reference to personal events (e.g. the marriage of a participant). Politeness in social conversation is for the benefit of the immediate interlocutors. In the environment of a press conference, which is recorded and may be televised, politeness may be required to accommodate the needs of a much wider range of listeners.

Partington lays great emphasis on the varying roles the participants play, in particular the podium speaker as motivator, spokesperson or relayer (speaking for himself, for the organization or reading a message prepared by someone else), and he writes of the 'absent motivators'. In his data and in the NATO conferences we can also consider the different audiences for the dialogue and how that affects the language used. The spokespeople are addressing the immediate listeners present, their own employer organizations, the media organizations which will make use of the transcripts, world public opinion and the local population, all of whom have different interests. The journalists have to consider other journalists and spokespeople present as well as their employers: Are they making enough effort to get the quotes for the news reports?

At the level of linguistic expression (since it is after all words that really matter in these studies), Partington identifies significant recurrent patterns in the way podium speakers reformulate their responses, use parallelism in structures and repetition. The spontaneity of much of the dialogue and the need to satisfy multiple audiences produce pressures on language production that makes conventional phraseology so important (in line with the views of Pawley and Syder (1983) on how fluent speech is generated).


Howarth 2002 examined the corpus for evidence of spontaneous versus scripted speech and focussed on the role that conventional phraseology plays: e.g. the density of collocations and idioms in a speaker's utterance, the use of lengthy repeated clusters ('in the course of our normal duties') and the recurrent use of internally significant, highly culture-specific language (over 130 instances of variants on 'church notices' to refer to daily announcements of practica1 information surrounding the conferences (e.g. bus trips laid on for journalists). Partington, in his US corpus, identified a parallel recurrent cultural expression: 'singing from the same songsheet/page'. In both cases these expressions can only be discovered through reading and serendipity, there being no mark-up with such semantic detail.

The question facing the analyst is what kind of search strategy to adopt that will reveal significant linguistic features of the discourse and help us understand what kind of communication a press conference is. The chief electronic tool used here was Wordsmith Tools (Scott 1999).

IV.1. Quantitative analysis of the questions and responses

Following Partington, two separate text files were extracted from a small sample of 10 of the 836 conferences (totalling 40,000 words). These files consisted of the questions from the floor (Q-file) and the responses (R-fiIe). These sub-corpora therefore excluded the spokespeople's opening reports, which could be regarded as scripted monologues and therefore quite different from the interactive sections of the conferences. These are the raw statistics:
Table 1: Comparative statistics for sample sub-corpora

                 tokens   types    ratio    number of   sentence
                                            sentences   length

10-text sample   40,139   4,544    11.32    1,358       29.54
Q-file           5,999    1,390    23.17    186         23.83
R-file           16,078   2,286    14.22    572         25 .O6
10 texts-(Q+R)   18,062

The opening statements take up 45% of the total conferences in this sample, while the Responses are 2.6 times longer than the Questions. There are sometimes more than one response to a single question, and indeed the question may be explicitly addressed to more than one spokesperson or may be in several parts. We see very differing type/token ratios between the Q and R files. The impression that this gives of greater repetition and recycling of chunks of text from the podium is confirmed in the cluster analysis, where all the lengthy recurrent strings are found in the responses: for example, the following six 9-word clusters were found, all from different texts and four different speakers:
   If our soldiers come across them in the course of our duties, then
   we will arrest them. The other thing you mentioned that was we that
   ve "must surely come across them in the course of our duties" Well,
   we also have to deem ... That is that we are determined to detain
   them if ve come across them in the course of our duties. Our
   ability to move our focus ... Our position is clear . If ve come
   across them in the course of our duties, we will detain them. If
   that is we have a resolute determination to detain them if ve come
   across them in the course of our duties and if the circumstances
   are ... When ve come across them in the course of our duties, we
   will detain them and hand ...

This phenomenon could indicate that, working close together with a common goal and listening to each other daily, these speakers come to absorb conventional language and there is a tendency towards linguistic convergence. In contrast, the questioners are more disparate, each with his or her own interests to pursue and less influenced by each other.

The full sample of ten texts, which include the opening statements, gives us a higher average sentence length, of nearly 30 words, confirming the suggestion that these contributions are scripted. The Response texts have a higher sentence length than the Question texts, indicating, perhaps, a greater fluency on the part of the spokespeople and that their responses are assisted by the scripts that they have prepared.

IV.2. Keywords

This is, in general, one of the most valuable tools provided by Wordsmith. It enables the analyst very rapidly to get a picture of the significant lexical content of the texts. Keywords are produced by the comparison between two wordlists, one from a large, reference corpus and the other from the smaller, specialized corpus. The keyword list from the NATO corpus contains are large number of subject-specific names of organizations etc. (their order in the complete list is given in the first column), which clearly tell us little about linguistic features of interest:
3       SFOR
6       OSCE
7       Srpska
8       Republika
9       Kosovo
10      NATO
11      IPTF
12      Herzegovina
14      Bosnia

Of more interest is the high frequency of personal pronouns, a very crude indication of the face-to-face nature of the discourse:
1      we
5      you
7      I

High up in the list are words connected with the conduct of the conferences:
17     thank
23     morning
26     know
36     think
41     question

The next category, occupying much of the central rank in the list, are the content words associated with the real subject matter under discussion:
22     representative
33     police
37     elections
45     agreement
46     forces
47     military
48     refugees
51     ambassador
52     situation
53     municipal
57     implementation

These are neutral items, not loaded in attitude from either side of the podium. We also have words like 'compliance', 'violation', 'exhumation' and 'crimes' (614 times, of which 443 occur in 'war crimes'), indicating problems arising in the day-to-day work of peacekeeping, but it is hard to use the keywords list to identify relationships between the speakers. indications of politeness are scattered through the upper reaches of the list: 'Gentlemen' (442 times), 'Ladies' (only 404), 'thank' (2586 and 'thank you' 2463) and 'sorry' (592). However, in searching for the distinctive tone and style of this genre of dialogue, keywords are of limited use.

IV.3. Reading the Q and R files

Reading, as suggested above, is possible and desirable in the study of a homogeneous set of texts of this sort, as it is possible to extract a manageable, representative sample from the total Corpus. Significant features of such discourse must be examined qualitatively, to answer such questions as: How interactive is it? How complex are the exchanges? Do respondents answer the question? 1s there aggression or confrontation on one or both sides? What causes aggressive behaviour?

Looking at the Q-file first (by far the smaller), we can confirm some of the features that Partington identified in the White House briefings. For example, there are both dual questions to one addressee, using such phrases as "the second part of my question" and dual questions to more than one: "a question for both of you", "could we get an answer from both of you about your concerns" or "For Kelly and Sheena ... ". 'The following suggests more clearly that journalists may attempt to manage the flow of talk: "You can decide who will give the answer but on the verification problem ... ", which indicates almost a reversal of the superficial power roles. This may be resented by the podium, but in general politeness is maintained when the questioner acknowledges a departure from the norm: "Two questions if I may change the subject".

While much of the dialogue in the NATO conferences concerns the day-to-day business of peacekeeping in the region, there are many references to the interna1 conduct of the briefings and externally to the outside world. The former is seen in this attempt to clarify who can answer what:
   <Q> Patrick do you only represent IPTF I mean in a way Alex
   represents the UN and Rizah and everybody, what about you?

Then we have a complaint about the way briefings are notified, giving a useful insight into the practicalities of the journalists' working lives:
   <(Chris Stephen--AFP)> <Q:> Do you think that you guys could agree
   on a time to hold this press conference? Certainly, it might give
   the international community a bit more faith in what you say if you
   could at least get a time when you could all be together in one
   place to talk to us. It can be 11:15, 11:30, 12, whenever you want.
   For instance, tomorrow, can we agree now on a time? Is it possible?

External references are generally to unspecified reports that the journalists are aware of: "there has been a lot of discussion in various forums this week", "Wendy, please. Media reports suggested recently that international humanitarian agencies officials in Kosovo ... ". In many cases the spokespeople know about these, but they may be at a disadvantage if the reports are in the local media: "It was reported in our local media and to Sarajevo also." In these cases the tables are turned again and the journalists may have more information than the speakers from the podium.

Other significant forms of questioning include the leading question:
   <Q:> Regarding the visit of Louise Arbour. Is there any comment
   from her or from you about this so-called trial of lzetbegovic
   taking place tomorrow in Banja Luka and will the tribunal, does it
   recognize, the trial? Presumably it doesn't.

It is not unusual for journalists to indicate a point of view rather than putting a neutral request for information. In the following example, we see a very explicit declaration of political attitude (worth quoting in full):
   <Q:> Now when we heard that there exists a eastern and western part
   of Republika Srpska, before 1 put my question to you, 1 would just
   like to make my statement. Wire tapping is one of police methods
   being conducted in every state. Particularly, regarding this
   matter, it is famous for being done by CIA in the United States.
   It's not just a secret that the CIA is tapping the President of the
   Limited States himself. My question would be referring to the right
   of the President to appoint a Chief of Police in Banja Luka. It is
   quite clear that in constitution of Republika Srpska, you were
   mentioning that constitution several times Mr. Haselock, it is
   quite clear that the President of the Republika does not have a
   right to perform duties which are in exclusive responsibility of
   the Minister of the Interior. By giving your support to the
   President in this duty, the international community is interfering
   in the interna1 affairs of Republika Srpska in a direct way, and it
   violates the constitution of the Republika Srpska. What would be
   your comment on this Mr. Haselock?

The response to this style of questioning is professional, but explicitly picks up the political slant:
   <Sirnon Haselock, OHR:> Firstly, you referred to my description of
   eastern Republika Srpska, and I detected from your remarks a sort
   of sensitivity that this may be in some way trying to indicate a
   regional split. East, west, south and north are geographical
   descriptions of parts of a country. One of your colleagues referred
   to the area, I think, Srpska Krajina. I live in the southwest of
   England, and some people refer to that part of the southwest of
   England as Somerset. So, there is no political significance in the
   geographical descriptions or names of parts of the country [...]

Here it is clear that there is a clash between the views of some factions of the local population, represented by a local journalist, and the official line of the international peacekeepers. It is perhaps this kind of evidence that makes this body of talk of such value. In addition to a political difference of perspective, there is also perhaps a cultural clash in styles of communication. The non-native speaker English of these conferences was commented on in Howarth 2002, and further examples can be found in this smaller sample:
   <Q:> what about the incident yesterday in the Office of the High
   Representative in Mostar, where some journalists was, I don't know,
   treated somewhere in the local press, there appeared yesterday and
   today. And the other question, what about this explosive near Celic
   and if the SFOR forces they are protected this area since its
   inter-entity borders. And what's the further measures you will take
   in this case?

In overall effectiveness in getting a message across native speakers are clearly at an advantage, and are able to defuse or dismiss expressions of local anger that are poorly delivered.

In addition to political sensitivity causing confrontation, there is a far more widespread level of aggression or irritation produced by professional frustration on the part of information-gatherers unable to do their job. The following selection gives a flavour. While some, such as the first, make accusations of political misconduct, the majority concern evasion: what Partington refers to as the distinction between a response and an answer.

<Q> You're here to defend democratic principals. Why don't you practice them in this case?

<Q> Is there any way I could get an answer today? I mean they're referring me to you and you guys are referring me to them and we're going around in a circle.

<Q> A question for Mr. Haselock, since it's my opinion that he didn't give me a precise answer on one part of my question.

<(Colin Soloway--US News)> <Q:> That's not a comment, that's a refusal to comment.

<Nancy Torner, UPI, Q:> O.K., so I won't get any answer on that, in terms of whether there really is any power of the international community to change such court cases.

While there are many clear indications of confrontation (e.g. finishing a question with "yes or no?"), there are problems in understanding the tone of a question from the transcript: "Alex, you said the IPTF would be monitoring ... " could be complaint or plain enquiry. The first is a request for clarification:
   <CRAIG:> General Jertz, some of this is just asking you to repeat
   what you said so 1 understand it more clearly but you said, if my
   notes are correct, that more than 300 tanks, artillery pieces, APCs
   and trucks had been struck.

The second, in a longer exchange, is more of an accusation that the speaker has gone back on his word:

<Q> But, I read that you said for Vecerne that case has been concluded now.

<R> No, I didn't say that.

<Q> So, you deny what you said for Vecerne?

<R> I would say they misquoted me. I said that we fully expect this case to be concluded soon.

<Q> Okay, thank you.

<R> Very soon.

<Q> Thank you

This takes us on to the R-file, consisting of the responses to the questions in the Q-file. There are certain parallel features. Where a journalist niay ask a dual question, there may be a coordinated response from the podium:

<a. ivanko:> no.

<Itc. m. marriner:, I do not either.

This co-ordination of responses may involve some light-hearted banter on the platform:

<Itc. m. marriner:> would you like to take this one?

<a. ivanko:> no!

<Itc. m marriner:> thanks!

Helpfully, the transcriber has considered this style of interaction significantly different from normal and inserted exclamation marks. At times the language of spokesperson and journalist meshes with a great deal of repetition of each other's phraseology (Notice, again, the exclamation marks to indicate the tone of exasperation.):

<q: Cjane bennett-powell--channel4 news)> who do you hope would use the service?

<k. janowski:> anybody can use it. the Serbs can use it, Croats can use it and Muslims can use it. whoever wants to use it. we have no idea who is going to use it. we have not done it yet and we hope both sides are going to use it, all three sides are going to use it, in this particular instance.

<q: Cjacques charmelot--afp)> Chris, again on the same subject; is there any registration procedure or do people just get in the bus and ride along?

<k. janowski? there is no procedure. it is like a bus, it is for free, they just get on the bus and ride on it!

<q: Cjacques charmelot--afp)> is it a free ride or you have to pay fee on it, buy ticket?

<k. janowski>: no, it is a free ride!

The podium speakers also make reference to externa1 press reports, in the following case to provide an opportunity to clarify the situation, which seems to be well managed:

<Itc.m. marriner:> I know this came <sic> raised its head a little bit last week and it was brought up again in a CNN and Reuters interview on Sunday, and I think senator Mitchell brought it up again yesterday, 1 do not know if the office of high rep wants to add something to that?

<c. murphy:> basically it is a good question but the answer is a bit premature. we are not there yet. in the interview that Kurt and CNN did on Sunday that question was discussed. and I think the question that you put to Mr. Bildt was "is not this a thermonuclear weapon in the high rep's arsenal?" in a sense it is, yes. and if I recall the interview question properly, Mr. Bildt was asked would he actually use such a weapon and he said that the whole deterrence effect depends on demonstrating ones ability to use it. so the answer is, yes of course, sanctions could be re-imposed, but we are not at that stage, that is something further down the line.

Evasion clearly exists, and the conventional ways of doing this are used, sometimes blankly refusing or dismissing the question, at others denying knowledge or responsibility, but often with an explanation and politely, even helpfully, making suggestions:

<Johan Verheyden, OSCE:> On TV Mostar, this is a MEC issue, it is being discussed within the MEC, and that's as far as I'm willing to comment on that.

<Aisling Byrne, OHR:> Yeah, well, we've dealt with that 7 days ago.

<ltc. marriner:> I cannot give you any details at the moment, because we have to make sure we go through correct notification procedures.

<P. Svensson:> I can't speak on behalf of the ICTY, but you should try and pose that question directly to Justice Arbour tomorrow evening. That's my suggestion.

<Wendy Rappeport, UNHCR:> My most recent information is not from the ground--it is from media reports. So, I can't help you with that one. But, if you want to call me this afternoon, I can get a-hold of the Geneva office and get a feel.

Irritation and defensiveness is also apparent from the podium:

<Simon Haselock, OHR:> Are you suggesting that we are going to recommend that we take similar action with Serb Radio and Television?

<Simon Haselock, OHR:> What are you suggesting? What similar action are you suggesting? What do you mean by similar action?

<Simon Haselock, OHR:> Firstly, I just answered the question about whether Madanie Plavsic asked us to carry out the action in Banja Luka, and I said she didn't. [...]

The response "are you suggesting" could be a neutral request for clarification from the questioner, but here the exchange seems to be more confrontational and the journalist becomes defensive in turn:

<Q:> I just wanted to ask a counter question?

<Q:> I'm not suggesting anything, I'm asking a question.

IV.4. Probing search strings

The close reading of these separate Q and R files has provided a useful insight into the significant language of the two groups of speakers. The most productive next step for finding significant data is an iterative one, using both maiiual and automated techniques to build up a clearer picture of the discourse. A number of key phrases have been identified and used for experimental concordance searches of the whole Corpus. These can be seen as probes that are sent into the data and may or may not extract useful examples. They may well produce results different from those expected. These will now be reported.

IV.4.1. "good question"

This is a conventional response (occurring 57 times), which may be a form of evasion or buying of time. In five instances the comment is followed by "but ... "

<R> basically it is a good question but the answer is a bit premature.

<R> and do they get a lawyer? it is a good question. I can't answer that, I don't know, I will check.

<R> good question. I have failed you.

IV.4.2. "what I'm saying is"

Partington talks of reformulation as an important aspect of the interaction. In the NATO Corpus, there appear to be a few different types using the same expression.

1. a defensive re-formulation usually following 'no':

<Maj. Boudreau:> We determine whether or not that would be used as part of an Investigation.

<Q:> are you telling us that for the time being, it's a part of investigation by whom?

<Maj. Boudreau? no, what I'm saying is, we're not prepared to discuss the release of that cassette at this point ...

2. a speaker being careful about the exact context:

<Ariane Quentier, UNHCR,> Then, you make your own investigation and you demonstrate scientifically that what I'm saying is wrong.

3. a reformulation after an explicit request for clarification:

<Q> In other words, he was saying that Milosevic has not seen fit to withdraw a single soldier sofar, despite 4,400 airstrikes. What are you saying now to clarify this?

<GENERAL MARANI:> What I'm saying is that if Milosevic wants to put more soldiers in Kosovo, that is his decision ...

IV.4.3. "there have been reports"

This indicates a 3rd party source and may show that journalists have more information than the peacekeeping organizations:

<Q> There have been reports in the media.

<Q> Again, there have been reports in the German press.

<Q> There have been reports in Reuters, and I think also in AFP.

<Q> On the Radio St. John issue, there have been reports in the local press.

<Q> there have been reports in the Serb press.

<R> a statement issued by our Brcko office last evening. It goes as follows: "There have been reports in some Federation media on 11 and 12 February, referring to incidents of unrest in Brcko town and the surrounding RS territory".

IV.4.4. negative questions

Reading a few examples in the transcripts suggests that negative questioning generally indicates confrontation, yet the picture is rather more complex. To take one example, why * n't you, shows that there are several discourse uses of this range of questions.

1. asking about action:

<Q> If you're concerned about force protection, why aren't you training your people on how to deal with this ... ?

2. complaining about quality of replies:

<Q> That doesn't answer my question. Why don't you admit that you don't know that you hit the hospital, you are still not sure?

3. The same question form may be returned from the podium:

<Q:> Why don't you know how much was taken?

<Alexander Nitzsche, OSCE:> Why don't you patient yourself and wait until these investigations are finished?

<SECRETARY GENERAL:> Why don't you help me a little bit and don't ask me to qualify, the word "permissive"? I think you understand what I mean.

4. The form is not always confrontational:

<Q:> Since there's not much else going on, why don't you tell us the whole story? <(laughter)>

<Alex Ivanko:> <(laughs)> O.K. Saturday, just before 3 p.m., the Iranian Ambassador to Croatia was travel ing on the Arizona road towards Tuzla ...

IV.4.5. "you guys"

The use above by a journalist of "you guys" shows a degree of informality that might suggest aggression. On the part of a podium speaker, however, it seems generally used to express friendliness.

<R> Thank-you. Questions? No more statements from anyone, believe it or not. Questions? Any questions?

<R> Come on. 1 didn't drive down here all the way, for no questions, come on you guys.

<R> Since no questions, we will see you tomorrow.

IV.4.6. Culture-specific references

Although the peacekeeping organizations and their representatives are multi-national, there is a strong bias towards US and British cultural allusions throughout these conferences, and it seems that there is a convergence effect as non-native speakers adopt these references into their daily speech. This was illustrated in Howarth 2002, with the example of a large number of instances of 'parish/church announcements/notices', and Partington uses the example of 'singing from the same page/songsheet'. We find an extended variation on this as well as another highly specific allusion:

<R> The international community is united on this issue and singing from the same song-sheet. Even if some of the voices are alto and some soprano

<R> Simon put it very well when he said that its not over until the fat lady sings. What that really means in real life is that it is not over until the OSCE says it is over.

<R> As I am an expert on using cliches, it is not over until the fat lady sings, and that may be some time yet

<R> But as we say where I come from, it's not over until the fat lady sings.

This last utterance is a clear example of speakers from the podium having a strong sense of place and origin and that they are addressing a (partly or largely foreign) multi-national audience. Sometimes these references help to provide personal background, at others to suggest a moral as well as geographical distance:

<R> You know there is a famous proverb, or saying, where I come from which says "we will weather the weather, whatever the weather"

<R> Next, we have been asked by an institute back in Canada--Nova Scotia, where 1 come from--a place called the Pearson Peacekeeping Center--- ...

<R> So, we'll let you know just as soon as we know, but as we say where I come from, it's not over until the fat lady sings.

<R> We can't do that in the United States, where 1 come from ...

<R> We are faced with that moral issue, that these are men that go in and beat up women, I mean, and where 1 come from that's considered very cowardly


The results of this study point to two main conclusions. Firstly, the methodology used to examine such relatively small and homogeneous texts should involve close reading of a sizable sample. The discourse features of interest cannot be revealed by automated, statistical means, unless there were a great deal of mark-up, which would entail even more laborious reading of the Corpus. This two-level approach allows for serendipity to throw up much useful data. There is also a value in looking separately at the question and response files, even though the discourse is essentially interactive. It is then possible to get a better understanding of the styles of each side as well as how they influence each other.

Secondly, the analysis has revealed in more detail the characteristics of the dialogue, many of which confirm Partington's findings, and in many ways it is shown to be more complex than might be assumed. For example, the power relationship is not all one way. The media can be seen to take some control over how the briefings are conducted and are at times very demanding. This can be identified through certain conventional indicators of aggression and confrontation. However, there is also informality, friendliness and humour on both sides. Overall, the impression is given of a convergence of language towards conventional phraseology that is picked up and spread among all participants.


Howarth, P. (1996). Phraseologv in English Academic Writing. Tubingen. Max Niemeyer Verlag.

Howarth, P. (2002). The conventional language of press conferences: scripted and spontaneous English speech. In Nuccorini, S. (Ed.) Phrases and Phraseologv--Data and Descriptions (pp. 127- 144). Bern. Peter Lang.

Levinson, S. (1 983). Pragmatics. Cambridge. Cambridge University Press.

Partington, A. (2003). The Linguistics of Political Argument. London. Routledge.

Pawley, A. & Syder, F. (1983). Two puzzles for linguistic theory: nativelike selection and nativelikefluency. In Richards, J. & Schmidt, R. (Eds.), Language and Cottittiztnication (pp. 19 1-225). London. Longman.

Scott, M. (1 999). Wordsmith Tools. Version 3. Oxford. Oxford University Press. Stein, G & Quirk, R. (1991). On having a look in a Corpus. In K. Aijmer and B. Altenberg (Eds.), English Corpiu Linguistics (pp. 197-203). London. Longman.

Stenstrorn, A-B. (1994). An Introduction to Spoken Interaction. London. Longman.

Peter Howarth *

University of Leeds

* Address for correspondence: Peter Howarth. The Language Centre, University of Leeds, LEEDS LS2 YJT, United Kingdom, tel: (44) (O)\1 13 343 3250, fax: (44) (0)1 13 343 3252, email:
COPYRIGHT 2006 Servicio de Publicaciones de la Universidad de Murcia (Murcia University Press)
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2006 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Howarth, Peter
Publication:International Journal of English Studies
Article Type:Report
Date:Jan 1, 2006
Previous Article:Occasional adnominal idiom modification--a cognitive linguistic approach.
Next Article:Occasional phraseological synonymy.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters