The International IJSLL (print) issn 1748-8885

IJSLL (online) issn 1748-8893

Journal of
Speech, Article
Language
and the Law

Suspects’ resistance to constraining and

coercive questioning strategies in the
police interview

Phillip Newbury and Alison Johnson

Abstract
Discussions of institutional discourse have frequently focused on asymmetrical power
relationships between participants. However, although power and control have been
the focus of much critical writing in discourse studies, far less attention has been paid
to resistance as an interactional resource. This article uses a qualitative Hallidayan
approach to explore questions and responses in an interview between police and Dr
Harold Shipman, focusing on resistance strategies employed by him in relation to the
institutionally more powerful interviewer. The interviewee actively resists powerful
questions through contest, correction, avoidance and refusal, whilst negotiating the
costs of resistance, such as challenge to the consensus of control and flouting of the
Cooperative Principle (Grice 1975) that would result in presenting an uncooperative
speaker. The interview is seen as a site of resistant struggle, where agreement and
disagreement are actively negotiated and where each participant attempts to produce
different goals and opposing roles.
Keywords: resistance, questioning strategies, police interview, institutional
control, constrain, coerce

Affiliations
Phillip Newbury: University of Aston, UK
Alison Johnson: University of Leeds, UK
Correspondence: Alison Johnson, The University of Leeds, School of English, Leeds LS2 9JT
email: [email protected]

Acknowledgement
The authors would like to thank the unnamed referees and Professor Malcolm Coulthard for comments and
suggestions on drafts of this article.

IJSLL vol 13.2 2006 213–240 doi : 10.1558/ijsll.2006.13.2.213

©2006, equinox publishing LONDON
214 The international journal of speech, language and the law

1 Introduction
Discussions of institutional discourse have frequently focused on asymmetrical
power relationships between participants: in courtrooms (Drew and Heritage
1992, Harris 1984, Eades 1994), in police interviews (Heydon 2005, Fairclough
1995), in classrooms (Fairclough 1989, 1995) and in a wide range of social life
(Fowler et al. 1979).
However, although power and control have been the focus of much critical
writing in discourse studies, far less attention has been paid to resistance as an
interactional response and resource, with the exception of Drew (1990, 1992),
who examines various ways that the witness can resist the lawyer’s questioning
strategies in courtroom cross-examination, Fairclough (1995: 50–2), who shows
how a youth suspect is non-cooperative when interviewed by a police officer,
and Clayman (2001), who examines interviewees’ overt and covert resistance
to question content in broadcast news interviews.
This article analyses an extract from a single interview (presented in full
in the Appendix), and focuses on the resistance strategies employed by an
interviewee in relation to the institutionally more powerful police interviewer.
The interviewee, an English GP Dr Harold Shipman, is being questioned about
the suspicious death of a female patient in his care. Shipman was subsequently
found guilty of murdering 15 of his patients by injecting them with an overdose
of morphine, although a Department of Health investigation carried out sub-
sequently into the deaths of other of his patients estimated that between 1974
and 1998 he had probably murdered 236 patients. (For more information on
the Shipman case, see Summers and the BBC link in the bibliography.)
In this study resistance is regarded as an active speaker choice, an act of
identity within a social constructionist view (Potter and Wetherell 1987), which
sees identity as being enacted discursively and dynamically. The interviewee
is shown to actively resist powerful moves by the police questioner in four
ways: through contest, correction, avoidance and refusal. Resistance clearly
has potentially significant costs in this setting, since doing resistance presents a
challenge to the consensus of power and control that expects cooperation. It is in
the interviewee’s interests to present himself as a cooperative interlocutor, since
resistance could imply guilt. The interviewee therefore has to adopt strategies of
resistance that do not constitute outright flouting of the Cooperative Principle
(Grice 1975) and that are acceptable within the rules and roles constituted by
the institutional interviewing situation.
There is an extremely healthy body of literature that discusses how, in legal
questioning, the interrogative form the questioner uses is designed to constrain
the respondent’s reply in different ways (Maley 1994, 2000, Conley and O’Barr
1998, Ehrlich 2002, Hill 2003, Gibbons 2003). For example, the questioner
 Suspects’ resistance to constraining and coercive questioning 215

will often use ‘yes/no’ interrogatives that constrain the respondent to confirm
or deny the presupposition contained in the question (such as ‘you murdered
her, didn’t you?’). Also, the questioner may use ‘WH-word’ interrogatives that
constrain the respondent to ‘filling in a blank’ about some aspect of the presup-
position contained in the question (such as the oft-quoted example ‘when did
you stop beating your wife?’). However, this strategy does not always produce
the expected or preferred response and, in such cases, resistance occurs. This
paper analyses the activity of resistance within the constraint frames embedded
in the police interviewer’s interrogatives, employing a Hallidayan framework to
qualitatively analyse the extract. The Shipman data is particularly interesting,
because it represents a struggle for control through use of the interrogative
form and resistance to that control within a broadly cooperative, but at times
resistant, response frame. The following extract serves as an example (P = Police
interviewer, S = Shipman).
25 P: There’s certain facts I need to make you aware of at this stage.
I don’t think there can be any of dispute in a lot of them.
Mrs Mellor’s body was buried on the 18th May 1998 at Highfield
Cemetery, Stockport. Would you accept that from me?
26 S: If you say.

P’s question is formed as a polar interrogative, expecting a ‘yes’ response.

However, intentionally or not it is a complicated question, because Shipman
might in ‘accepting’ be heard to be committing himself to having known the fact
about the burial before the policeman stated it. Thus the pair of turns produces
a cooperatively congruent relationship between question and response, whilst
omitting the ambiguous token ‘yes’. Thus resistance is enacted by omission of the
expected token, whilst at the same time appearing cooperative. Shipman signals
that a consensus of control is not absolute, as he resists the constraining polar
interrogative, and this resistance continues in the subsequent interaction.
This strategy is part of the activity of producing a context that allows for
qualification rather than absolutes and is a tactic in the interviewee’s attempt to
maintain and construct an identity that is non-culpable in the face of an accusa-
tion of murder, which, though not addressed in this question, is part of the
overall goal-focused agenda of the interview. As Fowler (1991: 52) points out
in relation to consensus in news reporting, ‘contradiction is resolved by accom-
modating variety within consensus’: “we agree to differ” ’. Here Shipman seeks
some accommodation as he resists ‘dichotomization practices’ (Fowler 1991:
52) of ‘yes’ and ‘no’ and ‘I’ versus ‘you’, by drawing attention to them through
the marked use of ‘if you say’ and resisting ‘yes’. Apart from ‘yes’, the unmarked
response to the question in turn 26 would have been ‘I would’. Shipman is
cooperative in terms of not invoking his right to silence and agreeing to answer
questions, but he is clearly a resistant interviewee and therefore the strategies
he employs shed some light on the practices and nature of resistance.
216 The international journal of speech, language and the law

2 Methodology
A Hallidayan systemic functional approach to the data has the benefit of being
an activity-based model of language, which is best summarised by Butt et al.
(2000). They observe that the functional approach allows the analyst to:
…recognise that words have functions as well as class and that how a word
functions can tell us much more than any description of words in terms of
class can about the piece of language, where it occurs, the person who chose
to use it in that function, and the culture that surrounds the person and the
message. (Butt et al. 2000: 29)
In other words, the systemic functional approach distinguishes between form
and function in language, and recognises that the way an utterance functions is
linked to the context in which it occurs. (In the analysis below upper case is used
for Hallidayan terms for functions, such as Actor, Identifier and Material and
Relational Process.) This makes it an ideal framework for analysing language
in situations where the interlocutors’ roles are constrained – such as during
the police interview – because it can differentiate between the language form
a speaker is using and the job they are trying to accomplish. Thus, it can be a
useful framework for the analyst when ascertaining if a suspect is attempting
to resist the constraints a preceding question placed on their utterance. Indeed,
Drew (1992: 477) suggests the suitability of the Hallidayan framework for this
type of analysis, when he observes that, whilst the lawyer must ask questions
and the witness must provide responses during cross-examination, ‘Other
activities may be done in the context of ‘questioning’ and ‘answering,’ but those
other activities are done through the format of questions and answers’.
The following extract from the Shipman transcript serves to illustrate what
Drew means.
37) P: I’d like to put it to you, doctor, that you interrogative tag q
were the person who administered that lady accuse

You were the person who administered

that lady with the drug
Identified Relational Process Identifier

38) S: No. answer

reject

Here, the interviewing officer’s tag-interrogative form is simultaneously func-

tioning as an accusation and as a question, whilst the suspect’s negative is
simultaneously a rejection of that accusation and an answer. In our analysis
of types of resistance, we suggest that this is contestation by the interviewee,
 Suspects’ resistance to constraining and coercive questioning 217

since he answers ‘no’ when ‘yes’ is expected. A Hallidayan analysis of the

functions of the parts of the clause shows how Shipman is the one identified
with the act of administering the drug, or as Halliday (1985: 115) says, the
clause identifies him as ‘the only member of the class’ of possible people who
administered the drug.
There are, of course, links to other qualitative schools of discourse analysis
such as Critical Discourse Analysis (CDA), Conversational Analysis (CA),
discourse analysis and pragmatics. Butt et al. (2000: 275) go on to observe that a
systemic functional methodology enables a pragmatic study of language in use,
because the framework accounts for the meaning of an utterance by considering
the context in which it is used. However, it is important to consider the limita-
tions of qualitative methodologies. Coupland and Jaworski (1999) discuss the
disadvantages of discourse analysis and point out where these methodologies
require support:
Discourse analysis is a committedly qualitative orientation to linguistic and
social understanding. It inherits both the strengths and the weaknesses
associated with qualitative research. As weaknesses, there will always be
problems in justifying the selection of materials as research data … If
discourse analysis is able to generalise, it can normally only generalise about
process and not about distribution … The point is that qualitative, interpre-
tive studies of particular fragments of discourse are not self-sufficient. They
need support from other traditions of research, even quantitative surveying.
(Coupland and Jaworski 1999: 36)
In other words, a qualitative analysis cannot say anything about the frequency
or probability of a language phenomenon. It can only observe the existence of
such a phenomenon and consider its significance in the context of use in which
it occurs; the limitations of this study are such that a wider scale investigation
and quantitative survey of resistance has not been carried out. Nevertheless,
there are many benefits of a qualitative approach. Our objective is to report
a small scale but detailed study that may lead to or inform other research,
to isolate particular instances of resistance and to evaluate the effectiveness
of these strategies. We are not trying to identify the frequency or likelihood
that a suspect will resist questioning, but rather to consider the nature and
significance of resistance in a particular case. Neither are we limiting ourselves
or others to the four strategies identified, but believe that they form a basis for
further study.
218 The international journal of speech, language and the law

3 The functions of police questions

In a legal context, lawyers/barristers, or the interviewing officer in a police
interview, employ questions to perform either of the following functions:
There may be two objectives of legal questioning. One is a genuine process of
elicitation of information. Another is to obtain confirmation of a particular
version of events that the questioner has in mind. (Gibbons 2003: 95)
We can go further than this, however, and say that these two macro-functions
of information-seeking and confirmation-seeking questions can be further
divided into the following micro-functions. (This table summarises and devel-
ops Gibbons 2003: 95–112 and Maley 1994: 36.)

Table 1 Macro- and micro-functions of questions

Macro-functions
Confirmation-seeking questions Information-seeking questions

Micro-functions
Degree of Declarative plus tag as Open information-seeking Q Constraint
coercion request for agreement Specific WH-information Q increases
increases Tag question Either/or information Q ↓
↓ Bare declarative

To explain, questions that function to elicit information can be roughly classi-

fied according to the amount of information they request, whilst questions that
function to seek confirmation of a version of events can be further distinguished
by the extent to which they coerce the respondent to agree with the proposition
contained in the question.
In the next two sections we discuss confirmation and information-seeking
questions and consider their effect in terms of how they are resisted. Each is
further sub-divided according to the micro-functions identified in Table 1.

3.1 Questions that seek confirmation

These questions occur where the questioner constructs a version of events using
the declarative mood, and then asks the respondent to confirm or deny that
version of events. These questions seek ‘yes’ or ‘no’ as their expected answer,
and there is great pressure exerted on the respondent by the language of the
question and the context of the situation in which it is asked to provide that
minimal response, as Maley (1994) observes.
 Suspects’ resistance to constraining and coercive questioning 219

… the predicted or entailed answer [to confirmation seeking questions] is

either a confirmation or denial, i.e. yes or no. Clearly these are more likely
to constrain a witness’s reply than an information seeking question, i.e. a
‘WH’ question. Of the confirmation seeking questions, declaratives with
tags are the most constraining: ‘You saw him there, didn’t you?’ Or ‘You saw
him there. Is that correct?’ When witnesses attempt to go beyond a minimal
response they can be checked. (Maley 1994: 36)
Thus, as Maley notes, three question types can function to seek confirmation
of a represented reality:
• declaratives with tag, where the tag takes the form of the inverted
Finite/Subject form that contains a request for the respondent to agree
with a preceding proposition, e.g. turn 25: ‘There’s certain facts I need
to make you aware of … would you accept this?’;
• tag questions;
• bare declaratives that function as prompts for the respondent to con-
firm or deny an assertion made in them.
This section now examines examples of these questions that function to seek
confirmation in the Shipman transcript.

3.1.1 Declarative plus tag in the form of inverted finite/subject requests for the
respondent’s agreement
Turns 25, 27 and 29 in the interview transcript contain good examples of this
type of question.
25) P: There’s certain facts I need to make you aware of at this stage.
I don’t think there can be any of dispute in a lot of them.
Mrs Mellor’s body was buried on the 18th May 1998 at Highfield
Cemetery, Stockport. Would you accept that from me?
26) S: If you say.
27) P: Now would you also accept that the body of Mrs Mellor was exhumed
with consent of the Coroner on the 22nd of September this year?
28) S: If you say so.
29) P: And I think, from what you were saying earlier, you were aware
that a post-mortem examination was subsequently undertaken.
Certain samples were taken at that post-mortem for forensic
analysis. Would you accept this?
30) S: You’re telling the story, yes of course.

These questions closely resemble tag questions in that they follow the
declarative plus question form, but do not negate the Finite verb of a
Process in the way that a reverse polarity Mood tag does. (Compare turn
29 with: ‘You were aware that a post-mortem exam was undertaken and
220 The international journal of speech, language and the law

certain samples were taken for forensic analysis, weren’t you?’) Instead,
they are intended to prompt the respondent to make a brief comment on
the general validity of a proposition. Thus, like many tags they expect the
respondent to confirm the version of events presented in the question, but
they are less confrontational than the tag question and therefore designed
to minimise resistance. In this extract, the interviewing officer establishes
the facts of the case with a series of declaratives (as noticed by the suspect,
‘S’, in turn 30) in which he describes the various processes that happened
to the body of the deceased patient after death. These processes are rep-
resented in chronological order, with Circumstantial components such as
the adverb ‘subsequently’ serving to signal that order is being maintained.
Each statement has attached to it a question form (italicised) that has the
structure Finite + Subject + Residue, and functions to request S to confirm
the facts of the case as they are reiterated. (For example: would (finite)
you (subject) accept that from me (Residue)?) Thus, these question forms
have less of a coercive effect (of demanding confirmation or denial) than
tag questions (described in Section 3.1.2) and are used after narration of
generally commonly accepted details. Their coercive effect is diminished
because there is a degree of interpersonal modality, as in turn 25 in the
projected clause following the mental Process (‘I don’t think there can be
any dispute in a lot of them’). This modality hedges the certainty of some
of the following propositions, thus reducing the pressure on the suspect to
confirm or deny them and less need for resistance and defence. Also, the
modal Finite ‘would’ used with these question forms expresses a degree
of modality that softens the demand for the suspect to accept or reject the
officer’s propositions.
It is perhaps not surprising that the suspect is able to avoid answering
every question in this sequence, without undue consequences, since they
are not particularly contentious, although he still uses the sequence to signal
resistance to polarisation. In turn 28 Shipman manipulates the experiential
function of language to perform avoidance. By representing the interview-
ing officer as the Participant Actor involved in a polar verbal Process (If
you say so) in turns 26 and 28, Shipman is not answering for himself, but
making an observation on what the police officer is doing. A similar resist-
ance tactic is used in turn 30. By avoiding agreement with details that are
established and commonly accepted, Shipman recognises that they predict
a longer sequence of questions directed towards blaming him – which is in
fact the case here – and so he attempts to disrupt the interviewing officer’s
flow. Avoidance is therefore also delay. Atkinson and Drew (1979) observe
that, in courtroom cross-examination, witnesses are sometimes able to
recognise that the lawyer is building towards a damaging accusation based
 Suspects’ resistance to constraining and coercive questioning 221

on the question they are being asked and a judgement of its relationship with
earlier questions, and so attempt to defuse the lawyer’s line of questioning in
what CA calls the ‘pre-sequence’ (Atkinson and Drew 1979: 142). Avoidance
can therefore be seen as a deflection strategy that attempts to cut off the
sequence before it can build to the more damaging accusation.

3.1.2 Tag questions

A tag is specifically intended to prompt a respondent to confirm or deny a
version of events presented in the question and is therefore more coercive than
the confirmation questions discussed in Section 3.1.1. In fact, Tsui (1992: 92)
observes that ‘the very construction of a tag question suggests that the speaker
has certain assumptions and is biased towards a certain answer’. The question
presupposes one version of events, expecting agreement from the respondent.
The following example (introduced in Section 2) is representative of the use of
tag questions in the interview.
37) P: I’d like to put it to you, doctor, that you were the person who
administered that lady with the drug, aren’t you?
38) S: No.

In this example (see also turns 79–80 and 86–87) the declarative preceding the
Mood tag expresses definite content through its interpersonal meanings because
there is no modal of probability present, as Halliday (1985) comments.
Note that in a statement the modality is an expression of the speaker’s
opinion … Note also that even a high value modal (‘certainly’, ‘always’) is less
determinate than a polar form: that’s certainly John is less certain than that’s
John; it always rains in summer is less invariable than it rains in summer.
(Halliday 1985: 86)
Thus, the interviewer in turn 37 above is representing the suspect as the Actor
(‘you’ ‘the person’) in the polar material Process of administering (‘adminis-
tered’) a lethal dose of a drug and hence (implicitly) of committing murder.
The tag then functions to prompt the suspect into replying. The certainty of the
(highly damaging) accusation in turn 37, followed by the tag, has the effect of
inviting the suspect to reply with the expected response – that is, confirmation
of the proposition contained in the question. However, in this case the suspect
denies the accusation and thus openly contests the proposition, demonstrating
legitimate resistance, since a suspect is expected to be truthful, and rejecting
the asserted truth of the presented proposition: that he administered a fatal
dose of morphine. However, in rejecting the accusation, he does not provide a
replacement version of his own, suggesting who else might have done so. He
222 The international journal of speech, language and the law

simply rejects the identification of himself as a murderer. Therefore, we will call

this resistance strategy ‘contest’.
On other occasions, one can observe instances where the suspect produces
more – or less – than the expected minimal response to a tag question. The
following extract displays two such instances.
110) P: You choose not to remember. It wasn’t too long ago in this
interview where you were explaining you’d been to see the family,
checked the computer record, and was telling them all about this
angina. Was it?
111) S: It’s a rhetorical question.
112) P: Quite correct though isn’t it?
113) S: I still have no recollection of entering that onto the computer.

By classifying the interviewing officer’s first tag question as ‘a rhetorical question’

the suspect is implying that it doesn’t need an answer, and hence is avoiding
having to produce one. (We see him employing a similar tactic in Q+A pair
120–121.) However, the police interviewer attempts to manage this resistance
and regain control of the interview by pressing the suspect for an answer in
turn 112, but in answering this second tag the suspect employs a similar tactic
to that observed by Drew (1992: 481) who notes that, in courtroom cross-
examination, the witness sometimes uses ‘I don’t remember’ in response to
a question from the lawyer when they want to avoid answering positively or
negatively. Also, by using this device, Drew (1992: 483) argues that the witness
is implying that the details they are being questioned about were not significant
enough to remember. Both turns (111 and 113) can be seen as resistant, since
the responses are ‘dispreferred’ (Pomerantz 1984: 64). In utilising the above
tactics, the suspect is producing what we will term ‘avoidance resistance’. The
most obvious advantage of not giving the expected ‘yes’ or ‘no’ answer, but
instead ‘something in-between’ – an ambiguous answer or avoiding giving any
answer at all – appears to be that the suspect cannot later be accused of lying.
At the same time, Shipman maintains his non-culpable identity, by providing
grounds for avoiding the questions: firstly, that the question is not a question
and then that he truly doesn’t remember at this point, resisting the officer’s
proposition in turn 110 that he chooses not to remember, thus also resisting
through correction.
In the final example, rather than implying that the question doesn’t need an
answer, Shipman asserts this, giving grounds for his resistance and avoidance.
120) P: … You told her you can remember she refused the treatment and
that’s what’s put on these false records isn’t it?
121) S: You see you’re stating the obvious, that doesn’t need an answer.

In this example Shipman practices what Clayman (2001) calls ‘overt resistance’,
whilst in the previous example the resistance is ‘covert’.
 Suspects’ resistance to constraining and coercive questioning 223

When resisting overtly, interviewees engage in various forms of ‘damage

control’. When resisting covertly, interviewees take steps to render the resist-
ance less conspicuous. Both sets of practices facilitate resistant responses by
reducing the negative consequences that might otherwise follow (Clayman
2001: 403)
The damage control in turn 121 is such that Shipman attempts to avoid any
negative inferences being drawn to his resistance, by drawing attention to the
officer’s ineffective framing of the question, whilst in turn 113 resistance is
rendered less conspicuous by using the ‘I don’t remember’ answer instead of
‘yes’ or ‘no’.

3.1.3 Bare declaratives that prompt the respondent to reply

Since interviews have a consistently alternating question and answer structure,
turns that are not formed as interrogatives are still interpreted as functioning
in this way. Nevertheless bare declaratives as questions are marked forms and
as marked declaratives they are the most coercive of the questions described
thus far. In turn 39 the interviewing officer uses a declarative structure that
represents through its experiential meanings the suspect as the Participant
Actor involved in the material Process of murdering the deceased patient
(italicised).
39) P: The levels were such that this woman actually died from toxicity
of morphine, not as you wrongly diagnosed – in plain speaking you
murdered her.
40) S: No.

This declarative contains no interpersonal modal verbs that would modify the
certainty of the interviewing officer’s proposition. It is an assertion that contains
no Mood tag, and thus indicates that the interviewing officer is making a state-
ment that he is very sure of, and the suspect is not even invited (overtly) to reply
(although at some stage the officer will require the suspect to admit their guilt).
It is the unmodalised, definite content of the declarative that functions as an
accusation that forces the suspect into employing an opposition strategy. The
officer presents the proposition that the suspect is guilty of murder as if it were
in the realm of commonly accepted detail – as though he expects the suspect
to accept its validity. This coercively provokes a response: acceptance or denial.
But this is not the only possible response. A third possibility is correction. This
is the case in turns 23 and 24.
23) P: From the progression you’ve noted on your records this was not
something that would have been expected.
24) S: This was a progressive angina. And people do die of coronaries
with no preceding history of angina.
224 The international journal of speech, language and the law

In this extract, the suspect produces a declarative correction, rather than the
agreement expected by the negative polarity of the policeman’s declarative
– interestingly in such cases the addressee can agree with either ‘yes’ (‘you
are right’) or ‘no’ (‘you wouldn’t’). Correction is therefore a covert resistance
strategy, since he avoids agreeing either way, but, since this is a dispreferred
response, he gives an explanation to make it less conspicuous. Drew (1990:
45) observes this phenomenon in courtroom cross-examination. He notes
that the witness might sometimes ‘other-repair’ the propositions made by the
lawyer, correcting although not necessarily challenging completely the version
of events proposed in their leading questions. By doing this, Drew (1990: 46)
postulates that the witness is correcting some negative implication embedded
in the lawyer’s question, in this case the negative implication that the death of
the patient does not correlate with the medical history recorded by Shipman.
Thus, he adds an addendum to the interviewer’s observation: that a coronary
can be unexpected.
In this section we have seen that questions that seek confirmation can be
resisted by three of the four strategies: contest, avoidance and correction.
We have seen that the degree of coercion increases with the type of question,
declaratives with tags being least coercive and bare declaratives most coercive.
Coercion is employed when resistance is likely to be high. Thus, the interview-
ing officer uses tag questions for details that there ‘can’t be much dispute about’
(to gloss his formulation in turn 25) and bare declaratives are used where the
content is most likely to be disputed. It appears, then, that the most assertive
propositions are contained in questions where the content is least certain and
the most provisional information is explored in the least assertive way. This is
paradoxical but logical, since posing questions the other way around would
consign those questions least likely to be disputed to the realm of serious
challenge and the most confrontational to the realm of least challenged and
therefore ease of resistance. Interrogation would therefore be rendered com-
pletely ineffective since nothing, not even most likely to be agreed upon details,
could be confirmed by the interview. In the next section we look at resistance
to information seeking questions.

3.2 Questions that seek information

This type of question occurs when the respondent is invited, to a lesser or
greater degree, to contribute information to the interview that may be new
to the questioner, and cannot therefore be confirmed. Nevertheless, despite
having a greater freedom to produce more than a yes/no response, the amount
of information the respondent can produce in a reply to these types of question
in a legal context is still constrained by the questioner, as noted by Gibbons.
 Suspects’ resistance to constraining and coercive questioning 225

The more information included in the question, the greater the questioner’s
control of the information, so the answerer can contribute less new informa-
tion. (Gibbons 2003: 101)
Three types of interrogative that function as information-seeking questions are:
Finite/Subject forms that contain open invitations to the respondent to supply
narrative, such as ‘would you like to tell me anything about – ’, WH- questions,
and either/or questions. Sections 3.2.1 to 3.2.3 discuss these three question
forms that function to elicit information in the Shipman transcript.

3.2.1 Finite/subject invitations to provide narrative

These questions are the least constraining, although the information contained
in the questioner’s utterance can direct the scope of that reply towards a specific
topic, or allow the response to be fairly open-ended, as in Gibbons’ (2003: 103)
example: ‘Is there anything that you would like to tell me in relation to this
matter that we are talking about?’.
Turns 31 and 32 constitute an example of this type of question and answer
pair.
31) P: A Home-Office pathologist – Dr Rutherford – carried out that
post-mortem examination. I think as you were going to mention,
his findings do not support that this lady died of a coronary
thrombosis as you diagnosed. Would you like to make any comment
on that – that finding?
32) S: Doctors don’t always diagnose a heart attack as a heart attack,
they’ll call it a coronary thrombosis or myocardial ischaemia or
myocardial infarction. To the average run of the mill GP they
are all the same – the patient’s dead. With a coronary thrombosis
you’d expect that there’d be a bit of heart that’s sort of dam-
aged but you can have had just an electrical disorganisation of
the heart which kills you just as effectively and leaves no symp-
toms at all – no signs, sorry, signs at all.

This question occurs at a point in the interview before the respondent is first
accused of murder, and thus the interpersonal modality present in the invita-
tion (‘Would you like – ’) could be seen as a way of softening the request for
information and putting the suspect at ease and ‘make any comment’ is also
relatively open, but is constrained in terms of commenting on the ‘finding’.
The presence of the median value modal ‘would’ is less coercive than if the
questioner had used an utterance containing a modal expressing a stronger
speaker opinion (Halliday 1985: 75) or a command, containing no modality,
to demand a reply from the respondent. In turn 31, the question (italicised)
invites the suspect to offer narrative on the specific topic of the pathologist’s
findings. In this sense, the only constraint that this question places on the
suspect’s answer is to remain focused on that topic. However, the questioner’s
226 The international journal of speech, language and the law

use of the adjective any expects a negative response. Heritage (2006) in a

study of physician questioning, ran an experimental trial on the question:
‘Are there ANY/SOME other issues you want to talk about today?’. He states
that ‘any’ ‘is negatively polarized: it ordinarily occurs in declarative sentences
that are negatively framed (e.g. ‘I haven’t got any samples’), and is normally
judged to be inappropriate in positively framed declarative sentences (e.g.
‘I’ve got any samples’)’. ‘Some’, on the other hand ‘is positively polarized: it
ordinarily occurs in positively framed declarative sentences (e.g. ‘I’ve got
some samples’), and is normally judged to be inappropriate in negatively
framed ones (e.g. ‘I haven’t got some samples’)’. For Heritage, this led to a
hypothesis that ‘although both ‘some’ and ‘any’ can appropriately be used
in questions, their polarized associations may have a direct causal influence
that biases responses’. His study tested for this effect and found that ‘among
those patients who had indicated more than one concern and who had not
yet discussed the additional concern(s) (N=98), the percentage of affirmative
responses was: some = 90% and any = 53%’. His conclusion was therefore that
‘the word ‘any’ is clearly negatively polarized: in nearly half the questions its
invitation to the recipient to answer ‘no’ overwhelms the opportunity afforded
by the question to mention additional concerns’. In the Shipman interview
the question, ‘Would you like to make any comment?’ is therefore communi-
catively ambiguous. On the one hand it invites Shipman to comment but on
the other it constrains, and the negative polarity of ‘any’ expects a negative
answer: ‘No, I wouldn’t like to comment’. However, given the narrative that
precedes the invitation, there are added implications for simply saying ‘no’.
Saying no, in these circumstances, would imply that Shipman doesn’t want to
comment because he doesn’t want to argue with the pathologist’s findings, but
if he doesn’t he implicates himself in the death of the patient. This question
therefore gives an opportunity to say ‘no’ but also challenges Shipman to
comment, with a warning. Any is pragmatically coded in the police interview
context since the interview begins with the caution, containing the word
anything: ‘You do not have to say anything. But it may harm your defence if
you do not mention when questioned something which you later rely on in
court’. Making any comment here is therefore part of the anything that doesn’t
have to be said, but can be seen as meaningful by omission by members of the
court, later. Shipman’s response is to employ a combination of avoidance and
correction here, avoiding commenting directly on the pathologist’s findings
and correcting the proposition, by going into a general explanation of the
problems faced by the medical profession in diagnosing a heart attack. His
use of the following is evidence of this generality:
 Suspects’ resistance to constraining and coercive questioning 227

• the non-specific plural noun doctors;

• the pronoun they;
• the non-specific noun group the average run of the mill GP.
His avoidance of the additional features is also evidence of resistance:
• avoidance of the word findings;
• avoidance of the pronoun I;
• avoidance of the name of the pathologist.
Shipman seems to employ these resistance tactics because the pathologist’s
report brings into question his own diagnosis, and so he is under pressure to
defend himself, a pressure that is underlined in the polarity of the question.
In being general in this way, he is able to reinforce his diagnosis and account
for the disagreement between himself and the pathologist because he suggests
in his reply that an indeterminate number of people in the medical profession
would similarly disagree on how to diagnose what ‘the average run of the mill
GP’ would recognise as a heart attack. At the same time he attempts to maintain
his non-culpable persona.

3.2.2 Wh- questions

Ehrlich (2002) explains that WH-word questions are not very constraining
on the respondent’s reply because they ‘impose little of the questioner’s inter-
pretation or words on the testimony: there is no proposition contained with
the WH-question other than the notion that “something happened” ’ (p. 196).
However, whereas the question form described in the previous section func-
tioned to invite the respondent to make any comment, WH- questions restrict
the respondent’s reply to specific aspects of the event in question, such as the
manner in which it was done (‘how…’ questions) or the time of occurrence of
the event (‘when…’ questions) as Halliday and Hasan (1976) observe.
A WH- question requires the specification of a particular item which is as
it were missing from the clause. The respondent…merely has to fill in the
blank. The WH- expression itself indicates whether the missing item is a
participant or circumstance, and various other things about it… (Halliday
and Hasan 1976: 210)
Thus, WH- questions are more controlling of the information that the respond-
ent provides in their reply than the open invitation to provide narrative.
However, after it is first asserted by the interviewing officer that the suspect
has committed murder, the suspect employs strategies to resist giving the
expected ‘filling in the blank’ answer to WH- interrogatives – as shown in the
following example:
228 The international journal of speech, language and the law

84) P: …You attended the house at 3 o’clock and that’s when you murdered
this lady and so much was your rush to get back, you went back to
the surgery and immediately started altering this lady’s medical
records. We can prove that only minutes after 3 o’clock on that
date you were fabricating that false medical history for this
woman. You tell me why you needed to do that?
85) S: There’s no answer.

In this example the WH- question is preceded by the command ‘You tell me’
(italicised). This coercive use of language, lacking in politeness markers or
modal verbs to accompany the imperative demand for the suspect to provide
information, produces the effect of reinforcing the power relationship between
questioner and suspect, overtly diminishing the suspect’s power as a result.
Furthermore, in this example (and in turn 122) the police questioner makes
some presuppositions in the declaratives immediately preceding the WH-
question that are incongruent with the suspect’s version of events, namely,
that the suspect is guilty of murder and that he falsified medical records to
cover up what he had done. The WH- form then functions to ask the suspect
to provide a reason why he performed these actions. In other words, the WH-
form functions here as a means of obtaining confession of the suspect’s role in
a presupposed version of events the questioner has in mind. Gibbons (2003)
comments on the use of presupposition in legal questions:
A technique for controlling information that is particularly clever or …
insidious (depending on one’s view) is to use presupposition in such a way
that the information content included by the questioner is made difficult to
challenge. (Gibbons 2003: 101, 105).
Thus, the presumptions made by the questioner, presented as definite declara-
tives, are difficult to deny without the suspect making an emphatic statement
not licensed by the question: a response that is not the expected one. In fact, the
suspect does not attempt to provide a reply to the why question (i.e. ‘because…’)
since to do so would mean confessing to the crime. Instead, by saying ‘There’s
no answer’, he is in essence openly refusing to answer. (See turns 81–82 for
another example of this.) This is different from the tactics listed previously,
because (unlike when he classified the interviewing officer’s interrogative as
a rhetorical question) the suspect does not recognise the question as invalid,
but simply declines to provide an answer to it. Thus, we will call this resistance
strategy ‘refusal’. This is clearly uncooperative in almost any context, yet it is not
entirely remarkable here. Suspects know that they cannot be forced to speak;
that is licensed by the caution and invoked when speech is likely to incriminate,
even at the expense of inferences being drawn.
 Suspects’ resistance to constraining and coercive questioning 229

3.2.3 Either/or questions

Of all the information-seeking questions, either/or forms are some of the

most controlling, since they restrict the respondent’s reply to making a
choice between one of two options presented as declaratives in the question.
One could regard either/or questions as a fusion between confirmation
and information-seeking questions, since although the questioner may not
know which of the options they offer in their question is the correct one,
they can effectively eliminate from the respondent’s reply any other version
of events. Turn 88 is an example of an either/or interrogative form that is
extremely constraining upon the respondent’s reply, because it additionally
employs the tag as a coercive confirmation seeking strategy. In this example
the first question is an either/or question that seeks information, but this is
followed up by a confirmation question. Both are resisted by overt avoid-
ance, with grounds being that the officer is merely telling a story and that he
should continue. This strategy by Shipman resists information giving whilst
encouraging the officer to give more information about the construction of
the police case or hypothesis.
88) P: You see if you examine that record which I’m going to go through
with you very shortly now to give you the exact time that things
were altered, it begs the question, did you alter it before you
left the surgery, which indicates what you’ve done was premedi-
tated and you were planning to murder this lady, or as soon as
you got back did you cover up your tracks and start altering this
lady’s medical records? Either way it’s not a good situation for
you doctor is it?
89) S: Continue the story.

In turn 88 the interviewing officer represents the suspect as the Participant

Actor involved in the two polar material Processes (‘did you alter … did you
cover up your tracks and start altering’) that describe the act of altering the
patient’s medical records, and thus he is implying that the suspect had good
reason to want to alter those records (because he was covering his tracks).
Thus the questioner presupposes that murder has been committed, and the
interrogative form presents only two, predetermined alternatives concerning
the tense of the action, that is the questioner wants to know when the suspect
did the falsifying of the medical history – before or after the act of murder.
Furthermore, the inclusion of the coercive tag functions to signal that the
interviewing officer is aware that the suspect is unable to make the either/or
choice without incriminating himself, because the tag expects a confirmatory
answer, which supersedes the either/or choice. Thus, here the interviewing
officer is pressing the suspect for a confession. Shipman again employs an
avoidance strategy to resist having to provide the expected answer to the
230 The international journal of speech, language and the law

officer’s question and hence avoids incriminating himself and admitting

his guilt. The imperative form ‘Continue the story’ functions to imply that
the interviewer is telling a story and not asking a question and therefore
no answer is required. Furthermore, the suspect’s retrospective labelling
(Francis 1994) of the officer’s prior speech as a story serves to imply that it
has no basis in reality.

4 Conclusion
In this paper we have outlined some of the resistance strategies that are
employed by one particular suspect in a single interview. As a result of detailed
analysis, we believe that the observations made here are generalisable to other
data, but recognise the need to test them on further data. This qualitative
analysis provides evidence of resistance strategies that are likely to be found
in other interview data and is not restricted to police interviews, and provides
a basis for further research. Using Gibbons’ (2003) dual typology of police
interview questions (confirmation and information seeking) we have shown
how a suspect can resist confirmation and information giving, whilst at the
same time managing reasonably plausibly to abide by the Cooperative Principle.
In addition, we have shown that, where cooperation is abandoned by refusal
to provide information or confirmation, the context sanctions this response
behaviour. In doing so the suspect is able to preserve his self-presentation of a
non-culpable identity and resist the questions’ functions: to represent him as
a culpable Actor in incriminating Material processes, such as administering a
lethal drug and murder. Thus the suspect resists transformation as an Actor
from doctor to murderer through opposition, correction, avoidance and refusal
of culpability in actions or Processes.
Our analysis has shown that suspects can assert their own power through
resistance. In turn 89, in response to an information seeking question,
Shipman uses the imperative: ‘Continue the story’, which is accepted as a
response. Resistance can be powerful. However, in spite of the performance
of power, many of the instances of resistance are simply rhetoric. They are
strategies to delay and deflect the reality of the evidence that the police
interviewer is revealing to Shipman and his solicitor. They resist acknowl-
edgement, coercion, culpability and confession and attempt to maintain and
preserve another identity. They are defensive moves in an interaction with
an inbuilt institutionalised challenge and defend role structure.
We have observed that, despite the constraining function of the inter-
rogative form, this suspect is sometimes able to employ strategies to resist
these constraints. Table 2 summarises the different strategies of resistance,
with some example utterances.
 Suspects’ resistance to constraining and coercive questioning 231

Table 2 Resistance strategies for information and confirmation-seeking (is and cs) questions

Strategy occurrence Example utterance

Contest (CS Q only) Answering ‘no’ when the question expects ‘yes’.

Correction (IS and CS Q) No. This happened/is the case.

Avoidance (IS and CS Q) I don’t remember.

It’s a rhetorical question.
Continue the story.

Refusal (IS Q only) [I have] nothing [to say].

There’s no answer.
Remaining silent.

These observations do not challenge in any way the notion that the interrogative
forms are used to constrain, control and coerce the reply a respondent can
give, but they do demonstrate that control can be resisted, even with the most
coercive questioning. Power and control are negotiated through interaction
where constraint and coercion produce a struggle, in which agreement, disa-
greement and resistance are produced. The interview is thus a site of resistant
struggle, where each participant attempts to accomplish different goals and
inhabits opposing roles: asking and demanding versus giving and resisting. For
the police interviewer, we see a struggle to control the suspect’s responses to
questioning, to coerce them into providing relevant and evidentially important
information in relation to their culpability, and to manage their resistance to
questioning. The interviewer seeks information, confirmation and responses
to evidence and hypotheses. For the suspect – especially one as skilful and
articulate as Shipman – we see a struggle to resist that control and to avoid
giving damaging information. The struggle is for identity, for Shipman. He
struggles to preserve his image as a caring doctor in the face of accusations
which claim to contradict that persona. As Raymond (2000: 355) points out
(in his analysis of reporters’ authority in live news broadcasts) authority is
something that speakers must continually achieve and sustain. Control is
therefore ‘tenuous’ because of the constant threat of challenge and resistance.
Resistance is therefore an oppositional and adversarial practice in response to
the ‘local accomplishment’ of police authority in the interview; it is in these
moments that interviewees assert their right to resist institutional control and
determine what is agreed and confirmed in the interview.
232 The international journal of speech, language and the law

Appendix

Key
P Police interviewer
S Dr. Shipman
C Shipman’s solicitor
Transcript of the tape of police interviewing Dr. Shipman about the death of
Winifred Mellor, taken from the Guardian Unlimited web site on 3 April 2004
(see bibliography).
1) P: Can I just hand you the medical records, again, of Winifred
Mellor. It’s exhibit AJB22 – that’s the folder there – if you
want to take the documents out – spread them out – so it’s
easier to refer to them in the best way you want. I’ll just move
that out of your way. OK. I want now to produce another exhibit
for you to look at, it’s a police reference OB24. Hand that to
you now. It’s the death certificate for Mrs Mellor. Off that
certificate will you just confirm for me what you attributed her
death to?
2) S: A coronary thrombosis.
3) P: Were there any other conditions?
4) S: Listed there? Nothing that was relevant to the cause of death.
5) P: Can you tell me, from whatever sources you use, how you came to
that conclusion exactly that it was a coronary thrombosis?
6) S: Well the history of angina that’s of increasing severity. We
have a relatively sudden death. Sudden in the sense that the
patient hasn’t made the effort to dial 99–9 – three nines,
sorry, three nines. Hasn’t staggered round to the neighbours
and that’s often the story with elderly patients that have
coronaries fairly rapidly.
7) P: Were there any signs on this lady’s body which indicated to you
that that was the cause of her death?
8) S: There are no signs on the body.
9) P: There are no signs?
10) S: No.
11) P: So you base this conclusion on the history you had of chest pain
and angina?
12) S: Yes.
13) P: Is that the sole thing that you based on?
14) S: Yes.
15) P: How satisfied did you feel about that decision?
16) S: At the time, very satisfied.
17) P: Has that changed now?
18) S: If I’d have had a post-mortem I understand from my solicitor that
there was some apharoma [phon.].
19) P: Well rather than discuss it with your solicitor…
20) S: But nothing else.
21) P: Can you indicate to me how severe the angina was at all?
22) S: The severity of angina doesn’t indicate the likelihood of a
coronary. I’m sure your medical experts would agree that. She
 Suspects’ resistance to constraining and coercive questioning 233

had no signs of congested heart failure, she only got a pain if

she rushed but it was lasting two to three minutes. How does one
grade angina? Severe angina – you don’t even get out of bed.
23) P: From the progression you’ve noted on your records this was not
something that would have been expected.
24) S: This was a progressive angina. And people do die of coronaries
with no preceding history of angina.
25) P: There’s certain facts I need to make you aware of at this stage.
I don’t think there can be any of dispute in a lot of them.
Mrs Mellor’s body was buried on the 18th May 1998 at Highfield
Cemetery, Stockport. Would you accept that from me?
26) S: If you say.
27) P: Now would you also accept that the body of Mrs Mellor was exhumed
with consent of the Coroner on the 22nd of September this year?
28) S: If you say so.
29) P: And I think, from what you were saying earlier, you were aware
that a post-mortem examination was subsequently undertaken.
Certain samples were taken at that post-mortem for forensic
analysis. Would you accept this?
30) S: You’re telling the story, yes of course.
31) P: A Home-Office pathologist – Dr Rutherford – carried out that
post-mortem examination. I think as you were going to mention,
his findings do not support that this lady died of a coronary
thrombosis as you diagnosed. Would you like to make any comment
on that – that finding?
32) S: Doctors don’t always diagnose a heart attack as a heart attack,
they’ll call it a coronary thrombosis or myocardial ischaemia or
myocardial infarction. To the average run of the mill GP they
are all the same – the patient’s dead. With a coronary thrombosis
you’d expect that there’d be a bit of heart that’s sort of dam-
aged but you can have had just an electrical disorganisation of
the heart which kills you just as effectively and leaves no symp-
toms at all – no signs, sorry, signs at all.
33) P: Well in his expert opinion there was nothing to support your
diagnosis is what I’m saying.
34) S: And he couldn’t rule out a disorganised electrical activity in
the heart.
35) P: Forensic examination of the samples taken, including muscle
tissue, at that post-mortem have been examined. These are the
samples taken from Mrs Mellor. And there’s certainly a high level
of morphine still contained in her body – a fatal level to be
precise. Can you account for that?
36) S: No.
37) P: I’d like to put it to you doctor that you were the person who
administered that lady with the drug aren’t you?
38) S: No.
39) P: The levels were such that this woman actually died from toxicity
of morphine, not as you wrongly diagnosed – in plain speaking you
murdered her.
40) S: No.
41) P: An expert has examined the drug this lady’s been prescribed, from
the records you looked to earlier within the exhibit – AJB22 –
and from those records nothing has been found which could account
for the morphine in her body. Can you suggest how it got there?
234 The international journal of speech, language and the law

42) S: No.
43) P: From your records, which you’ve had access to for some time now,
can you point out where the visits you made to Mrs Mellor are
indicated on the records?
44) S: Which visits are you talking about?
45) P: Well you said there was a visit in the morning.
46) S: No I said that she came to surgery. Says here at his practice.
47) P: Can you just show me where that is?
48) S: I thought we’d got this perfectly clear, the 11th of the 11th here.
49) P: So that’s on page nine and it’s the second entry 11.05.98
– angina pectoris. I don’t understand what these terms mean here.
Perhaps you could explain that for me, is this the right place
I’m looking at?
50) S: Yes that’s the right place you’re looking at. And I read that
record out to you on the previous tape and if you wish I’ll do it
again.
51) P: please.
52) S: Blood pressure’s 140 over 80. The heart sounds were normal, no
oedema, no sign of failure. Pain only if she rushes upstairs in
– and it’s in the throat and now in the arms. It lasts for pos-
sibly two or three minutes, it gets better all the time – meaning
when she stays still it goes away. Refuses treatment.
53) P: Now you told us you went back to Mrs Mellor’s home when you found
her dead at roughly quarter past six, I believe it was, where is
the record of that please?
54) S: I went back? You know I went, for the first time, to see her
dead.
55) P: Well now you’re saying it’s the first time before you said you
couldn’t remember, which is it?
56) S: Well you’re taking it that it’s a repeat visit, I’m correcting
you in saying I can’t remember visiting at 3 o’clock, so for me
this is the first visit.
57) P: Where’s the record of that visit?
58) S: The record is on the front of the record there – 11.05.98 coro-
nary thrombosis cause of death.
59) P: And then there was a further visit later on with the family, is
there a record of that?
60) S: No.
61) P: Would you not normally make a record of that?
62) S: No.
63) C: There is, sorry, just to clarify, there’s another entry on
11.05.98 Owen [sic.] dead
64) S: There that’s fine: ‘Called in to see her 6.15 – dead, coronary
thrombosis, daughter Linda present.’ I’m afraid both of us missed
that, second entry from the bottom.
65) P: So on this case, as well as making a written record on the front
of your manila folder you’ve also got computer entry?
66) S: Yes.
67) P: When were those notes made doctor?
68) S: They would have been made the next day, hence the 12th of the 5th
’98, the final entry, which says see my note nine. Because I cor-
rected the date for her death when I – the machine – entered up
the date, that I entered the entry which was 12th.
69) P: So you’ve corrected that to the 11th?
 Suspects’ resistance to constraining and coercive questioning 235

70) S: Yes. To say that she died on the 11th but the machine records the
entry as on the 12th.
71) P: So these would have been done the day after?
72) S: Yes I wouldn’t have gone back to surgery just to enter it up on
the computer.
73) P: One feature of these statements from the family is that they
couldn’t believe their own mother had chest pains, angina, and
hadn’t been informed.
74) S: …by whom?
75) P: By her.
76) S: By her, thank you.
77) P: They also found it very hard to believe that she would refuse any
treatment she was given in relation to this diagnosis and I think
now we can answer why that was because she didn’t have a history
of chest complaints and heart disease and angina did she doctor?
78) S: If it’s written on the records then she had the history and
therefore…
79) P: The simple truth is you fabricated her history to cover what
you’ve done, you’ve murdered her and made up a history of angina
and chest pains so you could issue a death certificate and pla-
cate this poor woman’s family, didn’t you?
80) S: No.
81) P: We’ve got a statement from a detective sergeant John Ashley who
works in the field of computers. He’s made a thorough examination
of your computer, doctor, and the medical records contained on
it. First of all he’s produced a medical summary report which is
that document – I’ll put it on the table in front of me – there’s
a police reference of JFA41. And, in essence, it’s what you’ve
been looking at already off this lady’s medical records. A copy
of the records as they stand on the computer. Well because this
man’s an expert he’s able to interrogate computers and he’s gone
into this computer of yours in some depth and what he’s found is
that there are a number of entries that have been incorrectly
placed on this record to falsely mislead and to indicate this
woman had a history of angina and chest pains. What have you got
to say about that doctor?
82) S: Nothing.
83) C: Can I have a look at them?
84) P: I’ll go through them shortly and you’ll be able to see them. You
attended the house at 3 o’clock and that’s when you murdered this
lady and so much was your rush to get back, you went back to the
surgery and immediately started altering this lady’s medical
records. We can prove that only minutes after 3 o’clock on that
date you were fabricating that false medical history for this
woman. You tell me why you needed to do that?
85) S: There’s no answer.
86) P: Well there is. There’s a very clear answer: because you’d been to
her house, rolled her sleeve up, administered morphine, killed
her and you were covering up what you were doing. That’s what
happened isn’t it doctor?
87) S: No.
88) P: You see if you examine that record which I’m going to go through
with you very shortly now to give you the exact time that things
were altered, it begs the question, did you alter it before you
236 The international journal of speech, language and the law

left the surgery, which indicates what you’ve done was premedi-
tated and you were planning to murder this lady, or as soon as
you got back did you cover up your tracks and start altering this
lady’s medical records? Either way it’s not a good situation for
you doctor is it?
89) S: Continue the story.
90) P: I mentioned the first print out, if you want to refer to the copy
that came from exhibit AJB22, all this is a full print with the
record you have there doctor. But the officer’s investigations
went far further than that. I’ll just remind you of the date of
this lady’s death – 11th May ’98 – perhaps you can explain to me
then at three minutes and 39 seconds after 3 o’clock that after-
noon you have endorsed the computer with a date of the 1st October
’97 which is 10 months prior: ‘chest pains’. Please look at your
record there and look at the entry dated 1st August ’97, would you
do that for me please?
91) S: I’ve looked.
92) P: What does it say?
93) S: It says: ‘chest pain’.
94) P: 1st of August ’97. Or it starts off on that date, irrigation of
external auditory canal for removal of wax hair, this practice,
one ear OK.
95) C: What date do you say that was?
96) P: First, this is the 1st August ’97.
97) C: Yeah.
98) P: I’m referring to the part that’s been inserted, I’m now reading
what was on there already.
99) C: Right.
100) P: First, 8.97 seen in GP’s surgery here, this practice, Dr HF
Shipman. First, 8.97: ‘chest pain, here, this practice, all
appears OK, question mark, angina’. Which was what you made great
comment on earlier, saying ‘well look this lady’s got angina
– not unexpected’.
101) C: Sorry which bit of that do you say was added? All of it or just
the last bit?
102) P: Chest pain.
103) C: Just the last bit, so there are two entries you’re saying?
104) P: I’m now showing you – I’ll put it in the middle of the room
– so your solicitor can examine it as well then. It’s an exhibit
JFA42. And it’s an insertion, behind your computer there’s a
ghost image and it records what’s placed in, when and what’s
removed. I appreciate you’re writing, when you’re done I’ll show
you exactly together with your client. This record of information
was created it on the 11.05.98 this lady’s death, three minutes
39 seconds after 3 o’clock by HFS – Dr HF Shipman: ‘term, chest
pain, all appears OK, angina’, date: 01.08.97. And it was created
on 11.05.98. Where’s that information come from doctor?
105) C: Can I just have a look at that?
106) P: For the benefit of the tape I’ve handed the exhibit to Mr
Shipman’s legal representative. Mr Shipman is now looking at the
record himself. Thank you. I’ll ask you again doctor, where’s
that information come from?
107) S: I’ve no recollection of me putting that on the machine.
108) P: It’s your pass code, It’s your name.
109) S: It doesn’t alter the fact that I can’t remember doing it.
 Suspects’ resistance to constraining and coercive questioning 237

110) P: You choose not to remember. It wasn’t too long ago in this
interview where you were explaining you’d been to see the family,
checked the computer record, and was telling them all about this
angina. Was it?
111) S: It’s a rhetorical question.
112) P: Quite correct though isn’t it?
113) S: I still have no recollection of entering that onto the computer.
114) P: You’d been to see this lady at 3 o’clock on the same day as her
death. And you come back and alter the computer, so much so that
later that evening you’re telling the family exactly – ‘Look:
chest pains. 1st of August I saw her.’ There’s three people heard
you talking about it. It doesn’t stop there, it goes further than
that. I’ve got an exhibit here JFA43. If you want to look at your
record there doctor, this was made at the same date – the lady’s
death – 11.05.98 and was put on 1505 and 21 seconds. What are you
talking? – two minutes later than that first entry – 11th May.
I’ve got it on your record doctor. 26th of the first ’98, purport-
ing to have been made: ‘chest pain, odd times and exercise does
not let it stop her, angina, refuses test and RX’ – I presume is
referral – ‘not bad enough’. How do you explain that doctor?
115) S: Again in the same manner as I’ve explained the other one, I
cannot remember putting that on the computer. I’m well aware
that’s how an audit trail works.
116) P: It’s been placed on your computer on the same day that you’ve
seen this patient.
117) S: Yes.
118) P: Hasn’t it?
119) S: Well there’s no argument about that.
120) P: And you were able to tell the family all about these entries
later that evening at six o’clock. So it seems unusual to me
somebody sneaked in your surgery, updated these false records in
that three hour time span. Used your pass word, put your name on
it and you’re able to say: ‘Look, I told her, she wouldn’t take
treatment’ – that’s what that’s saying there. You told her you
can remember she refused the treatment and that’s what’s put on
these false records isn’t it?
121) S: You see you’re stating the obvious, that doesn’t need an answer.
122) P: Well it is obvious and I think it will be obvious why you’ve done
it, don’t you? It doesn’t stop there. 11.05.98, 1506 and 49 sec-
onds by HFS – Dr HF Shipman – a minute after the entry purporting
to come from the 26th of the first ’98. 01.05.98: ‘Mrs Mellor seen
here, this practice, term – ill, comment – very vague pain in
right [inaudible word], still not wishing tests or RX to let us
know if worse’. It’s no wonder the family didn’t know, Mrs Mellor
didn’t know did she? She had no angina and that’s why it came as
a complete shock to everybody. You told us at the start of this
interview doctor, didn’t you, you tried to update your records as
soon as possible, ‘as soon as practicable’ I think was the phrase
you used. You tell me why the visit you say you had at your sur-
gery from this lady on the date was not recorded until the day
after, yet all this false history is?
123) S: I’m sorry what have you just said? That I recorded, what did I
record the day after she died?
124) P: 12th of the fifth ’98, 0738, well I presume that’s very early in
the morning for you isn’t it doctor? I mean your surgery doesn’t
238 The international journal of speech, language and the law

start till half eight, does it, quarter to nine I think it was?
It’s reported on the 12th, 11.05.98: ‘angina pectoris’ – and a
part on your record here about the 140 over 80 – ‘no oedema, no
chest – is it calf pain – only if rushes up stairs in throat, no
in arms, lasted two to three…’
125) S: When are you saying I put this on the machine?
126) P: I’m not saying anything.
127) S: When is it being…
128) P: …computer records saying that it was recorded at 0738 41 seconds
on the 12th of the fifth. The computer has a built in clock and
times when the record was updated so you put these false records
on and then…
129) S: You’ll find the clock is out by an hour because we don’t change
it for summertime.
130) P: Which way?
131) S: So that fact would have been put on at 8.38 the following morning.
132) P: So all these other records would have been at 1606 and would have
been at 1605 and would have been at 1603 – is that what you’re
saying?
133) S: The timing seems a little awry, that’s what I’m saying because
the clock is not changed or we don’t change the clock for sum-
mertime.
134) P: I don’t think we should bother about clock timings or anything…
135) S: Well you’re making a great point at me being there at 7.30 in the
morning.
136) P: I’m making a great point about somebody falsifying a person’s
records on the date of her death to fabricate a medical history
for them.
137) C: Can we have a consultation at this stage please?
138) P: Certainly. The time now by my watch is 1712 hours and we’ll
switch off the tape.

Note
1 This article began as the sole work of Phill Newbury written and submit-
ted to the journal in 2005, during his first year as a doctoral student at the
University of Aston, Birmingham, UK. The Editors returned the manuscript
with detailed comments and suggested that he revise and resubmit. Tragi-
cally, Phill died very suddenly and totally unexpectedly in April 2006, before
he had made much progress with the revision.
The roots of the article lie in an assignment written for a Forensic Linguistics
module taken when he was a Master’s student at the University of Birming-
ham. The editors decided to invite Alison Johnson to revise and complete it
for publication as she had been involved since its inception. She had taught
part of the FL module, had watched the student paper develop and extend,
and had commented on several versions of the article before it was submit-
ted to the journal and after it was returned.
 Suspects’ resistance to constraining and coercive questioning 239

References
Atkinson, J. M. and Drew, P. (1979) Order in Court: The Organisation of Verbal Interaction
in Judicial Settings. Basingstoke: Macmillan.
Butt, D., Fahey, R., Feez, S., Spinks, S. and Yallop, C. (2000) Using Functional Grammar:
An Explorer’s Guide. (2 nd ed.) National Centre for English Language Teaching and
Research. Sydney: Macquarie University.
Clayman, S. E. (2001) ‘Answers and evasions’, Language in Society 30(3): 403–42.
Conley, J. and O’Barr, W. (1998) Just Words: Law, Language and Power. Chicago:
University of Chicago Press.
Coupland, N. and Jaworski, A. (1999) ‘Introduction: perspectives on discourse analysis’, in
N. Coupland and A. Jaworski (eds) The Discourse Reader. London: Routledge, 1–44.
Drew, P. (1990) ‘Strategies in the contest between lawyer and witness in cross-examina-
tion’, in J. Levi and A. Walker (eds) Language in the Judicial Process. London: Plenum
Press, 39–64.
Drew, P. (1992) ‘Contested evidence in courtroom cross-examination: the case of a trial
for rape’, in P. Drew and J. Heritage (eds) Talk at Work: Interaction in Institutional
Settings. Cambridge: Cambridge University Press, 470–520.
Drew, P. and Heritage, J. (eds) (1992) Talk at Work: Interaction in Institutional Settings.
Cambridge: Cambridge University Press.
Eades, D. (1994) ‘A case of communicative clash: Aboriginal English and the legal system’,
in J. Gibbons (ed.) Language and the Law. Harlow: Longman, 234–64.
Ehrlich, S. (2002) ‘(Re)Contextualizing complainants’ accounts of sexual assault’, Forensic
Linguistics: The International Journal of Speech, Language and the Law, 9(2): 193–212.
Fairclough, N. (1995) Critical Discourse Analysis. The Critical Study of Language. London:
Longman.
Fairclough, N. (1989) Language and Power. London: Longman.
Fowler, R. (1991) Language in the News. London: Routledge.
Fowler, R., Hodge, G., Kress, G. and Trew, T. (eds) (1979) Language and Control. London:
Routledge and Kegan Paul.
Francis, G. (1994) ‘Labelling discourse: an aspect of nominal-group lexical cohesion’, in
M. Coulthard (ed.) Advances in Written Text Analysis. London: Routledge, 83–101.
Gibbons, J. (2003) Forensic Linguistics: An Introduction to Language in the Justice System.
Oxford: Blackwell.
Grice, H. P. (1975) ‘Logic and conversation’, in P. Cole and J. L. Morgan (eds) Syntax and
Semantics 3: Speech Acts. New York: Academic Press, 41–58.
Guardian (2004) Transcript. http: //www.guardian.co.uk Accessed 3 April 2004.
Halliday, M. A. K. (1985) An Introduction to Functional Grammar. London: Edward
Arnold.
Halliday, M. A. K. and Hasan, R. (1976) Cohesion in English. Essex: Longman.
240 The international journal of speech, language and the law

Harris, S. (1984) ‘Questions as a mode of control in magistrates’ courts’, International

Journal of the Sociology of Language, 49: 5–27.
Heritage, J. C. (2006) ‘Reducing patients’ unmet concerns in primary care: a trial of two
question designs’. Paper presented to the American Sociological Association Meetings,
Montreal, Canada.
Heydon, G. (2005) The Language of Police Interviewing. A Critical Analysis. Basingstoke:
Palgrave Macmillan.
Hill, M. (2003) ‘Identifying the source of critical details in confessions’, The International
Journal of Speech, Language and the Law: Forensic Linguistics, 10(1): 23–61.
Maley, Y. (1994) ‘The language of the law’, in J. Gibbons (ed.) Language and the Law.
Essex: Longman, 11–50.
Maley, Y. (2000) ‘The case of the long-nosed potoroo: the framing and construction of
expert witness testimony’, in S. Sarangi and M. Coulthard (eds) Discourse and Social
Life. Essex: Longman, 246–69.
Pomerantz, A. (1984) ‘Agreeing and disagreeing with assessments: some features of
preferred/dispreferred turn-shapes’, in J. M. Atkinson and J. Heritage (eds) Structures
of Social Action: Studies in Conversation Analysis. Cambridge: Cambridge University
Press, 79–112.
Potter, J. and Wetherell, M. (1987) Discourse and Social Psychology: Beyond Attitudes and
Behaviour. London: Sage.
Raymond, G. (2000) ‘The voice of authority: the local accomplishment of authoritative
discourse in live news broadcasts’, Discourse Studies, 2(3): 354–79.
Summers, C. ‘Crime case closed infamous criminals: Harold Shipman’, BBC News Online,
http://www.bbc.co.uk/crime/caseclosed/shipman.shtml Accessed 3 April 2004.
Tsui, A. (1992) ‘A functional description of questions’, in M. Coulthard (ed.) Advances in
Spoken Discourse Analysis. London: Routledge, 89–110.