© The Author(s) 2019
James MurphyThe Discursive Construction of Blamehttps://doi.org/10.1057/978-1-137-50722-8_5

5. The (Non-)Assigning of Blame

James Murphy1  
(1)
Bristol Centre for Linguistics, University of the West of England, Bristol, UK
 
 
James Murphy
Email: j.murphy@uwe.ac.uk

5.1 Introduction

In this chapter, I will explore the formal end to the inquiry proper—namely its reporting. In particular, I will explore how, despite not being an overt part of the Terms of Reference of an inquiry, the reporting stage engages in blaming or blaming-like behaviours. In order to do this, I have compiled a corpus of inquiry reports from the 25 most recent inquiries first discussed in Chapter 2. In the previous chapter, I explored what blame is and gave an account of how witnesses may pre-emptively seek to avoid receiving it. In this chapter, we shall see whether or not such attempts are successful. As well as looking at how blame is done, I will also explore the discussions which some inquiry chairs engage in around blame—their metapragmatic commentary on it—and why they believe that blame is a risky action and one which should be avoided. I will offer a counter-narrative to this, one in which I suggest that blame at public inquiries is simply inevitable and something which should be embraced. I shall make a related argument, that blame is an important tool which ensures that there are consequences to negative actions; this is important for the smooth running of a civil society.

In considering how blame is done through inquiry reports, I have at the forefront of my thinking the following quote from Edwards and Potter (1992):

People treat each other, and often treat groups, as entities with desires, motivations, institutional allegiances and biases, and they display these concerns in their reports and attributional inferences. Anyone who produces a version of something that happened in the past, or who develops a stretch of talk that places blame on someone or some category of persons, does so at the risk of having their claims discounted as the consequence of stake or interest. [...] Because their discourse displays these concerns, we suggest that participants should be thought of as caught in a dilemma of stake or interest: how to produce accounts which attend to interests without being undermined as interested. [...] people can perform attributional actions such as blamings indirectly or implicitly through providing an ostensibly disinterested factual report which allows others to follow through the upshot or implications the report. (Edwards and Potter 1992:158, their emphases)

Whilst it might seem that an inquiry chair is free from the risk of seeming as though they have a stake or an interest, particularly given the legitimation work which was discussed in Sect. 2.​5, what will emerge from this chapter is that these concerns remain. They remain, in part, I would suggest because the chair is concerned that their report may be viewed as a whitewash and their hard work be dismissed in this way. In a sense, this increases the pressure on them to engage in blaming. But, this potential desire is mitigated by the problems associated with blaming which will also be discussed—not least the view that explicit blame can create a culture in which problems are covered up for fear of one’s reputation being damaged by being (publicly) held to account. In light of this, the core of this chapter will show how blame is discursively constructed, but rarely done explicitly. The inquiry chair seeks to strike a fine balance meaning they are not seen as a patsy of the establishment, whilst not engaging in blame which is over the top and runs the risk of contributing to a blame culture which is viewed as potentially damaging.

In the following Sect. 5.2 I will outline the general approach taken in analysing blame in the inquiry reports. In Sect. 5.3, I will discuss the case of the Shipman Inquiry report which is unusual in that it contains a section entitled ‘Who was to blame?’. This gives us an insight into the language used to blame. I discuss how blame can be considered as a speech act. In Sect. 5.4, I seek to explore whether the language associated with blame found in the Shipman Inquiry reports is replicated in other inquiries. In particular, I focus on the lexical field of blame. In Sect. 5.5, I discuss modality—the use of expressions which demonstrate a speaker’s attitude towards a proposition—and how such expressions might be employed in the performance of blame. I discuss views of blame presented by inquiry chairs in Sect. 5.6 and how such impressions may inform our understanding of the choices made by a panel when it comes to blaming (or not) in their reports. I present some views on why blame should be embraced in that section and in the conclusion (Sect. 5.7).

5.2 Approach

Taking the most recent inquiries to have been instituted, as discussed in Chapter 2, I built a corpus of the reports which had been published at the time the preliminary work for this chapter began. This gave a corpus of 23 inquiry reports, i.e. all those listed in Appendix A, except: The Independent Inquiry into Child Sex Abuse, which is not complete and has hearings scheduled at least until March 2019 and the Robert Hamill Inquiry, whose report has been completed (indeed it was finished on 25th February 2011) but has not yet been published as to do so may risk the right to a fair trial for those who may potentially be prosecuted for involvement in the murder of Robert Hamill.

The 23 reports vary in length—from around 80,000 word for the Bichard Inquiry to approximately 2.5 million words for the Iraq Inquiry. In total, this gives a corpus of just over 11.5 million words.

This corpus was interrogated for recurrent patterns, and was compared to language found in other political settings by comparing results from the public inquiry reports corpus with the Hansard Corpus. This corpus contains all contributions made in the Houses of Parliament from 1803 to 2005, approximately 1.6 billion words. The choice of such a comparison will be motivated later.

5.3 Explicit Blaming: The Shipman Inquiry

I have said throughout this work that blaming is not something which is done explicitly and have talked in a number of cases of covert blaming. In the introduction, I suggested that most of this chapter will explore how blame is done in an implicit way and suggested that I would be considering how the panel draw on the component parts of blame which were outlined in the previous chapter.

Whilst this is true for the most part, I have kept from the reader the case of the Shipman Inquiry which merits this section on ‘explicit’ blaming. Volume 2 of the Shipman Inquiry contains a most unusual section entitled: ‘Who was to blame?’ From that we can learn a great deal about the language used to do blaming. In turn, we can think about the linguistic means which are used to discuss the constituent components of blame; this will give us a stronger idea of the lexical or semantic field of blaming, which I will discuss in more detail in the next section. Moreover, looking closely at how blame is discursively constructed (e.g. the patterns of language which fall outside the blame semantic field) when the author has made clear that what she is doing is blaming, will allow us to see whether these patterns are also found in inquiry reports where blame is not made explicit in this way.

The advantages of this approach are hopefully clear—whilst exceptional, this relatively short section of an inquiry report can be analysed closely in a way that the large corpus cannot. It will also generate ideas which can then be methodically investigated in the corpus of inquiry reports.

5.3.1 About the Shipman Inquiry

The Shipman Inquiry was established to investigate a variety of matters in relation to the serial killer Harold Shipman, a General Practitioner who used his role as a community doctor to forge the wills of patients and subsequently murder them. The Inquiry, chaired by Dame Janet Smith (then a High Court judge) took place between 2000 (following Shipman’s imprisonment) and 2005. Its report came in six parts, each focusing on a different part of the inquiry’s Terms of Reference:

 
First report: Death disguised

explored how many patients Shipman killed and the means by which the murders were carried out.1

Second report: The police investigation of March 1998

considered why an earlier investigation by the police into the suspiciously high death rate among Shipman’s patients did not result in his arrest.

Third report: Death Certification and the Investigation of Deaths by Coroners

was an exploration into the procedures for investigating death and made recommendations to tighten up coronial procedures (particularly relating to cremation).

Fourth report: The regulation of controlled drugs in the community

outlined the systems in place for managing access to controlled drugs, such as those that Shipman used to murder his patients, and made further recommendations about this.

Fifth report: Safeguarding patients: Lessons from the Past—Proposals for the Future

explored how complaints against GPs are investigated and the practices of the General Medical Council in the face of allegations of wrongdoing against those who they validate.

Sixth report: Shipman—The Final Report

considered historic allegations against Shipman and the possibility that he began murdering patients when he worked as a junior doctor at Pontefract General Infirmary.

 

5.3.2 Who Is Blamed?

In the section ‘Who was to blame?’ in Chapter 16 of volume 2, Dame Janet Smith names three people who she views as being to blame for the police investigation having failed to catch Dr. Shipman earlier in his campaign of murder and prevent further loss of life. They are:

 
Chief Superintendent Sykes:

The officer to whom the coroner first reported the concerns of a doctor who had signed what was considered to be a high number of cremation certificates referred by Dr. Shipman. Sykes appointed one of his reportees, Detective Inspector Smith, to investigate the concerns of the coroner and the reporting doctor.

Detective Inspector Smith:

The investigating officer, who was supervised by CS Sykes. Smith was relatively inexperienced at conducting investigations.

Dr. Banks:

The doctor who was tasked with exploring whether the medical records of 14 patients would be sufficient to explain the cause of death listed on their death certificates. Banks found that two or three deaths may have merited further investigation by the coroner; this conclusion missed matters which were subsequently viewed as glaringly obvious (such as the patterns of visits from Dr. Shipman just prior to the patients’ deaths).

 
The blamees are listed in order of the amount of blame which Dame Janet believes they should carry—from most to least. This suggests that the view I took in the previous chapter, that having multiple people responsible for a negative action still can result in blame, was the right one. Indeed, ranking blame in the way that the chair does here also supports the view that I came to that blame should be viewed on a scale or continuum, rather than as a binary (either being blamed or not). This is, perhaps, most clearly evidenced by the chair’s comments about the role that Dr. Banks had to play:
../images/375431_1_En_5_Chapter/375431_1_En_5_Figa_HTML.gif

5.3.3 Lexical Items Involved in Blaming

In assigning blame to these three parties, the lexical items which do much of this work can be categorised in the ways shown in Fig. 5.1.
../images/375431_1_En_5_Chapter/375431_1_En_5_Fig1_HTML.gif
Fig. 5.1

Lexical items used to carry out blame in Volume 2, Chapter 16 (‘Who was to blame?’) of the Shipman Inquiry (frequencies in brackets)

These lexical items make clear that the chair believes the inquiry participants’ conduct fell short to such an extent that it becomes blameable either because: they were unable to recognise an action was needed and this inability resulted in a negative (blameable) outcome (example 78); the action which they took was the wrong one and contributed to the negative consequences (example 79); or their failure to act at all led to a predictably negative result (example 80).
../images/375431_1_En_5_Chapter/375431_1_En_5_Figb_HTML.gif
../images/375431_1_En_5_Chapter/375431_1_En_5_Figc_HTML.gif
../images/375431_1_En_5_Chapter/375431_1_En_5_Figd_HTML.gif
These examples give a snapshot of the sort of work which the chair is engaged in in this section; spelling out in quite clear terms what she believes the negative actions, inactions or consequences were for which each party is blameable. The attribution of responsibility for these actions is embedded within both the structure of this section and within the descriptions of the acts listed. It appears in the structure since each individual is given a separate subsection under which are listed a number of points of blameable conduct—this serves to highlight that these actions are ‘owned by’ these individuals. Moreover, the use of possessive marking in ‘his decision’ (79) and ‘his inaction’ (80) may clear where the chair believes the ‘ownership’ of responsibility lies. There are, however, a few occasions when the notion of responsibility is directly considered:
../images/375431_1_En_5_Chapter/375431_1_En_5_Fige_HTML.gif

Here, again, we find the idea that responsibility (along with blame) is something which can be ‘ranked’ or placed on a continuum. Whilst denying the primary responsibility for the negative action, the chair does not absolve DI Smith of responsibility fully (this example continues and a significant chunk of the rest of this paragraph can be found just previously in example 80). The implicature which Dame Janet’s statement carries is that DI Smith bears (some) secondary responsibility for the failure of the investigation. This is one of the few instances where responsibility is assigned through linguistic means—responsibility is otherwise inferred from the fact that the blamee carried out (or failed to carry out) an action.

The final lexical aspect which merits discussion relates to the notion of criticism. Whilst less frequent than giving an account of the blameable, the chair explicitly framing her action as criticising the blamee does occur.
../images/375431_1_En_5_Chapter/375431_1_En_5_Figf_HTML.gif

In stating that she is about to be critical of still other matters, the above example suggests that Dame Janet wishes that what she has said before about CS Sykes as also being critical. In so doing, she adds another element to blame which should be considered when looking at the wider sample of inquiry reports—that of criticism, i.e. actively expressing a moral judgement on the basis of the negative actions attributed to an individual or individuals.

The purpose of this subsection has been to collect together a set of lexical items which are used by an inquiry chair when explicitly outlining who was to blame. Their use in other reports will point towards blame being done, even if this is not explicitly stated in the report.

5.3.4 The Avoidance of ‘Blame’ as a Performative

In outlining the lexical items which are involved in doing blame, the absence of blame itself was hopefully noticeable. Except in the title of the section, the word blame occurs only once (and this will be discussed more in Sect. 5.3.6). This is perhaps surprising, especially if we consider blame to be a speech act.  After providing some background to speech act theory, I hope to show why the absence of the word blame to ‘do’ blame should not be seen as all that surprising.

5.3.4.1 What Is a Speech Act?

Speech act (SA) theory, first outlined in Austin (1962) and developed by Searle (see in particular Searle 1969), starts from the idea that language can be and is used to do things—i.e. performing an utterance can make a tangible change in the world (cf. the title of Austin’s book How to do things with words). Austin makes an important distinction between the levels at which one is ‘doing’ something in producing an utterance.

 
Locutionary act:

  The (physical) act of producing an utterance with a determinate sense and reference.

Illocutionary act:

The ‘doing’ of the act associated with the locution. Illocutions are intentional; they are what the speaker intends to do in producing the locutionary act.

Perlocutionary act:

  The effects which are brought about on the interlocutor/third parties as a result of the utterance of the locution. Speakers have in mind the perlocutionary effect(s) they wish to produce as a result of their utterance, but these are not guaranteed to come about. In the case of blame, such perlocutionary effects might include: the blamee feeling bad; third parties knowing that the blamee was responsible for the negative action; etc.

 
The other key tenet of speech act theory came about as a reaction to logical positivism—a philosophical movement in the ascendancy when Austin was first formulating SA theory in the 1930s. Logical positivism asserted (in its most strident form, see here Ayer 1936; Carnap 1928) that the only utterances which are cognitively meaningful are those which are verifiable—that is, those which can be said to be either true or false. Austin (1962) sets out to show the untenability of this position, in particular by discussing a class of utterance which does not seem to fit into this true/false dichotomy. Consider the following examples:
../images/375431_1_En_5_Chapter/375431_1_En_5_Figg_HTML.gif

It is clear that we cannot talk about the truth or falsehood of these utterances—it would be odd for a third party to say “that’s untrue!” after any of them. What Austin suggests as a more important concept for these examples is ‘felicity’ and the circumstances that must be in place for an utterance like one of those in (83) to be ‘happy’ (to use Austin’s term) which are described as the felicity conditions. Thus for the sentencing act in (83c), a number of felicity conditions must be fulfilled in order for the utterance to make a change in the world (i.e. the guilty party having perform 15 years hard labour): the speaker must be an approved person (i.e. a judge); the utterance must take place in an approved setting; the hearer must have been found guilty prior to the statement, etc.

Austin sets out some categories for these conditions (see Austin 1962: 14ff.), but it is the later development of felicity conditions which has been taken up by most SA theorists and so it is these which I shall outline here. Searle’s felicity conditions are not merely ways in which a speech act can fail (as is the case with Austin’s) but are actually constitutive rules (Searle 1969). This means that the felicity conditions of a particular act are definitional for that act and can be used to distinguish acts from one another.

Searle’s felicity conditions fall out into four component conditions:

 
Propositional content:

The proposition that must be expressed in the utterance

Preparatory condition(s):

A state of affairs necessary for the illocutionary force to be expressed

Sincerity condition(s):

The psychological state S has towards the propositional content

Essential condition:

  The utterance must “count as” the desired illocution

 

Below I will outline Vanderveken’s (1990) proposals for what these conditions may look like for the act of blaming, and what this might tell us for how blame can be achieved.

The final aspect of SA theory which it is necessary for me to introduce at this point relates to the difference between direct and indirect speech acts. Here it seems fitting to fall back on a much-used example sentence:
../images/375431_1_En_5_Chapter/375431_1_En_5_Figh_HTML.gif

Usually this would be treated as a request illocution (speech act), intended to get a third party to pass the salt to the speaker. But the literal force of the utterance is to question the ability of the third party to perform that same action. Indeed this mismatch is often exploited by parents who believe themselves to be humorous; my father would always insist on responding to the question in 84 with “I could” while folding his arms with a broad grin on his face. So examples like these are described as indirect speech acts—they have a literal force which is secondary and a non-literal force which carries the intended primary meaning (i.e. it is this non-literal force which is the intended illocution). We will see the implications of this for seeking to blame a third party.

5.3.4.2 Blaming as a Speech Act

In the previous chapter, I provided an extensive discussion of the nature of blame and came to a settled view upon it. I will return briefly here to defining blame, since it will help in making clear what linguistic means are available to speakers in performing such an act. There has, to my knowledge, been only one previous suggestion for how to define blaming as a speech act. Vanderveken says of it:

To blame is to express disapproval with an explicit attachment of this disapproval (mode of achievement and propositional content condition) to someone, perhaps the hearer, for having done something judged to be bad (preparatory condition). (Vanderveken 1990: 217)

The view taken by Vanderveken alights on most of the aspects which I discussed in the previous chapter and is broadly in line with what I settled on: that blame is an act in which a speaker attributes responsibility to a third party for a negative action. Implicit in that is the idea that the blamer disapproves of the action, because they would otherwise be viewed as a hypocritical performer of blame, resulting in an insincere speech act.

Vanderveken also considers a distinction which I have not so far, the difference between blaming and reproving (in part, I have not made this distinction because ‘to reprove’ is such a low frequency lexical item). However, Vanderveken’s (1990) take on to reprove merits discussion:

The expressive use of ‘reprove’ is that of expressing strong disapproval (preparatory condition) with the intentional action of an agent (propositional content condition), and with the sincerity condition of reprobation. It is this latter that distinguishes reproval from blaming, in that one may blame someone for (for example) throwing the ball through the window without necessarily reproving him for it.

The first distinction is one of measure: blaming involves disapproval, reproving involves strong disapproval. In a sense, this sort of distinction has been taken care of by my considering blame on a continuum and so one would expect reproval to exist at the most potent end of the blame spectrum. But Vanderveken argues that the most important things which separates blame from reproval is that the latter requires the speaker to have the intention of ‘doing reprobation’. In other words, the act of reproving involves censuring a hearer for their behaviour. In part that distinction is made necessary by another aspect of the (non-)performance of speech acts which Vanderveken seeks to incorporate earlier in his work, the notion of the public vs. the private. On this matter, he says the following:

One must also distinguish between speech act verbs like ‘accuse’ which name illocutionary acts which can only be performed in public and those like ‘blame’ which can be performed in thought alone and in silent soliloquy. When a speech act is essentially directed at a hearer who is different from the speaker, the speaker must have the intention to communicate his intention to perform that act to the hearer. Consequently, that speech act requires a public performance. (Vanderveken 1990: 168, his emphases)

I agree with Vanderveken that blaming can be done privately, but if it is done privately then I do not think we can think of it as a communicative speech act. And so the distinction between reproval and blame is not really one which has to be drawn, because blame which happens internally is not a true speech act, therefore blame which happens publicly is synonymous with reproval. But even without this sort of thinking, the distinction between reproval and blame is not necessary for our purposes: blame as it is discursively constructed at public inquiries is necessarily a public act. It cannot be done in silent soliloquy—and even if an inquiry chair might privately blame a participant, that private judgement is immaterial to the public pronouncements which we can examine and analyse.

Putting these discussions to one side, the most important thing which I want to draw out of the felicity conditions which Vanderveken presents is the idea that anything which is said that fulfils the above conditions can be viewed as performing the speech act of blame. Unlike—say—a judge who can only perform the act of sentencing a criminal by saying ‘I sentence you to...’, a speaker performing the act of blaming does not need to say ‘I blame X for Y’ (where X is the blamee and Y is the blameable). So long as the effect of the utterance is to make clear that the blamee was responsible for an act which is bad and which one disapproves of, then the act of blame is performed.

I should add that I do not think there has to be a one-to-one correspondence between an utterance (or a sentence, or whatever one’s preferred unit of language is) and a speech act. A speech act, instead, can be viewed as something which is constructed across multiple utterances—this certainly seems to be the case with blame found at the Shipman Inquiry. Different sentences set out the negative action, the assignment of responsibility for that action and (potentially) the speaker’s disapproval of the person in light of that. It is through multiple utterances that blame can be discursively constructed—as such, the expression I blame X for Y is not required to ‘do’ blame.

That still leaves a question of why ‘blame’ is rarely found. My view of this will be developed through the rest of the chapter, but to pre-empt this I would say that there is a general preference for blame to be done implicitly (and the reasons for this will explored later in this chapter). Using the performative verb blame does not allow this general preference to be realised. This, however, is not unique to the speech act of blame—consider how rare it is in English for a speaker to say ‘I request you to do X’ or ‘I compliment you on X’ to carry out the acts of requesting and complimenting respectively.

5.3.5 Mitigation Work

We have so far seen that investigating this one section of an inquiry report which is explicitly carrying out blame is a good source for lexical items used to contribute to this action. But there are other moves contained in this section which are worthy of discussion and which are also likely to help investigate the attribution of blame on the wider scale of all 23 inquiry reports. One move which is striking is that, as well as blaming, the inquiry chair engages in mitigation work—seeking to explore the reasons which may help to account for this negative behaviour, and which potentially reduce blame. That the chair carries out mitigation is evidence, first and foremost, that she believes her words to be carrying out the action of blaming and secondly, that she views blame as a risky action (since doing so unabashedly may open her up to criticism).

The first element of mitigation which is presented in this case is one which is quite specific to this inquiry—it relates to the difficulty which some involved in the investigation had in considering that a medical professional, Shipman, was capable of killing his own patients.
../images/375431_1_En_5_Chapter/375431_1_En_5_Figi_HTML.gif

I highlight this example because it sees a (different) inquiry chair pick up a concern raised by Rebekah Brooks in the previous chapter—that the inquiry has the benefit of hindsight and may judge actions through this prism which was (obviously) not available at the time of the commission of the potential blameables. Returning to the Shipman Inquiry this was also a concern raised earlier in the report; the chair gave voice to the view of a family member whose mother was killed by Shipman that the inquiry had to be cautious not to judge witnesses negative purely with the benefit of hindsight (Vol. 2, Chapter 16, p. 14). So relying on the view that hindsight should not be relied upon may well be a canny blame avoidance tacit, because we can see that it is something which an inquiry chair (and inquiry participants) are cautious about. Indeed, it seems to act in mitigation of a blamee.

But mitigation does not simply take the form of trying to avoid judgements based on hindsight which is likely to be applicable for all potential blamees. Mitigation at the Shipman Inquiry also involves the individual circumstances of those participants who are blamed.
../images/375431_1_En_5_Chapter/375431_1_En_5_Figj_HTML.gif

In this case, the fact that a more junior officer was not supported by his superior reduces the blame which he receives: it mitigates the fact that he failed in his role as investigating officer, in part because he should not have been assigned the role in the first place (cf. example 81) and because, when he was, supervision was not offered as it should have been.

Mitigation can also be offered where a blamee seeks to remedy the damage done by their negative actions—in this regard, blame reduction is not just about the circumstances which may account for the commission of the wrongdoing, but also involve how the transgressor responds to their wrongdoing.
../images/375431_1_En_5_Chapter/375431_1_En_5_Figk_HTML.gif

In this case, CS Sykes’ recognition of his errors and the apology provided (which also entails a realisation that his behaviour was wrong) stands in his favour. I will discuss the role of apologising in a great deal more detail in the next chapter.

But whilst potential mitigation might be discussed by the inquiry chair, it can also be dismissed as carrying weight in excusing the blameable behaviour of one of the three participants. For instance, the limits of excusability are scrutinised by Dame Janet in the following example which involves Dr. Banks’ assessment of the medical records of Shipman’s patients and the failure (or refusal) to recognise that they pointed either to Shipman having killed his patients, or at the very least provided them with less than optimal care.
../images/375431_1_En_5_Chapter/375431_1_En_5_Figl_HTML.gif

Here the distinction is drawn between mitigation and an excuse—but again, I think we are talking about the measurability of blame. Mitigation reduces the blame potential somewhat, whereas in the inquiry chair’s conception excusing could reduce the blame potential in its entirety. That these discussions around mitigation take place give further evidence that blame is happening—this is something that can be explored in more detail in the full inquiry reports corpus.

5.3.6 Blame for Non-investigable Matters

A further aspect which must be mentioned at this juncture is the ability of the chair to blame someone for a matter which is not fundamental to the Terms of Reference of the inquiry, but relates more to the conduct of the inquiry itself. Consider the following example:
../images/375431_1_En_5_Chapter/375431_1_En_5_Figm_HTML.gif

Here, the chair states her obligation to chastise DI Smith for lying when his conduct and involvement in the Shipman investigation was scrutinised by the police. Whilst Smith’s behaviour after the investigation had no material effect on how that initial investigation was conducted, and could not change anything about how it was conducted, it has made the process of arriving at a clear understanding of what went wrong harder. But what is particularly noteworthy about Dame Janet’s statement here is that it acknowledges that there are different interpretations of the same events which may result in different views of blame and blameability—this mirrors what I discussed in the previous chapter about views of blame not being static and universal. It is interesting to note that the chair accepts that seeking to avoid blame is a normal thing to want to do (and presumably not, in and of itself, blameworthy). Where this desire to evade responsibility becomes blameworthy, is if the blame avoidance behaviour involves lying. All this having been said, blaming for this matter is not at the core of what the inquiry has investigated.

5.3.7 What Should Have Been

We have seen so far that certain lexical items are associated with blaming—they allow the chair to frame actions as negative, suggest responsibility for them and to explicitly criticise those responsible. We have also seen that there is a general avoidance of the performative blame. Other signs of the presence of blame include mitigation work—since this indicates that interpersonally sensitive work must be being engaged in. But there are still further markers which indicate that the chair views the conduct of the three blamees as having fallen short. Consider the following examples (I have added emphasis):
../images/375431_1_En_5_Chapter/375431_1_En_5_Fign_HTML.gif
../images/375431_1_En_5_Chapter/375431_1_En_5_Figo_HTML.gif

The chair does not simply focus on framing negative actions as failures, mistakes or faults and on providing descriptions of these actions as we have seen, but these examples highlight her discussion of what should have happened. Outlining what should have happened makes clear a contrast between the actual behaviour (or the failure to act) of the witness and what would have constituted appropriate, non-blameable (or even praiseworthy) conduct.

The use of the modal marker should or ought allows the chair to express what she ‘considers ‘right’ whether morally [...] or as a matter of expediency’ (Huddleston and Pullum 2002: 186). The use of the past time with these modal markers (i.e. talking about what should have happened), ‘usually impl[ies] that the event did not take place’ (Palmer 1988: 133).

In implying that the ‘correct’ or morally right action did not occur, a further implicature is generated: that this failure to do what is right is blameable. I arrive at this view precisely because of what Edwards and Potter (1992: 158) say about people providing an ostensibly disinterested factual report which give hearers (or readers) the opportunity to work through the implications of such a statement. If we think about Grice’s maxim of relation (Grice 1975: 45)—which says ‘be relevant’—it must be relevant for the chair to outline what was not done and which ought to have been done; I argue that it is relevant because she is (implicitly) criticising the witness for having failed to carry out this action.

Modality as expressed by should and ought to is important in the Shipman Inquiry section. I will show later in this chapter that other modality markers are also associated with implicit blame.

5.3.8 The Responsibility of Blaming

Before I move on to explore the larger corpus of 23 public inquiry reports, I want to note one last issue which the section in the Shipman Inquiry report raises. The following example demonstrates that blame is something which the chair treats seriously; she does not take her role as the assigner of blame lightly. She recognises the strong emotions which are aroused as a result of the public inquiry and the fact that blaming parties after the inquiry can be seen as ‘branding’ them as wrongdoers.
../images/375431_1_En_5_Chapter/375431_1_En_5_Figp_HTML.gif

By showing her sympathy, Dame Janet makes it clear that her blaming does not come vindictively, but instead comes because it is necessary for her to do so. This is important—the legitimacy of the blame could be undermined by a blamer doing so gleefully; it may seem less considered and stemming from an emotional, knee-jerk reaction, rather than a considered assessment of the evidence. Note, too, how the chair frames the blame as being something of ‘their [the three participants’ –JJM] own making’, placing fault for it squarely with them. But finally, there are further signs of mitigation for the three witnesses—firstly, that it was bad luck for them to have been caught up in Shipman’s criminality in the first place, and secondly, that there would have been others who would have acted similarly, or at least, in a blameworthy manner when faced with the same circumstances. We will see later in the chapter that these discussions about the power of blame are something which are not unique to this inquiry.

5.3.9 Summary

This section has explored one part of the Shipman Inquiry report—the only one in the whole Inquiry Reports Corpus which explicitly flags up that it is about blaming. This has allowed us to explore the linguistic patterns which emerge when an inquiry chair is engaged in blaming and serves as a useful basis on which to explore the assignment of blame across the full corpus. It will allow us to see whether these elements are present even when an inquiry report does not make it clear that it is engaging in blame. In other words, in the rest of this chapter I am keen to answer to question of whether these linguistic features permeate the other inquiry reports as a means of implicitly and discursively constructing blame.

5.4 Lexical Field of Blame

In exploring the short section of the Shipman Inquiry, I identified that the performative use of the verb blame is not found. Instead, we find a cluster of lexical items which are utilised by the inquiry chair to highlight the negative actions which are blameable, the responsibility of the blamee for those actions and, sometimes, an explicit expression of her disapproval or criticism of an individual who is being blamed. This cluster of expressions form part of what we might call the ‘lexical field of blame’—terms which when used contribute to the discursive construction of blame. In this section, I will explore the frequency with which these expressions are used in the 23 inquiry reports. Of course the frequency alone may not reveal a great deal, and so I will also draw on another corpus to shed light on what the relative frequencies of these terms can tell us about the role of blame in inquiry reporting. In particular, I use the Hansard Corpus Alexander and Davies (2015), made up of oral contributions produced in the UK Parliament between 1803 and 2005. The speeches made in the UK Parliament are perhaps the closest genre that we have when it comes to a readily available corpus. Since parliamentary contributions are also likely to broach issues raised at public inquiries then this also seems like a reasonable comparison to draw. Seeing the relative frequency of blame expressions will give us a benchmark to see whether they are more or less frequent in the public inquiry reports.

Whilst the initial exploration of the Shipman Corpus was useful for giving an idea of the sorts of expressions which can be used to perform blame, it clear that this small sample cannot provide us with all of the expressions which are in the ‘blame repertoire’ of inquiry chairs. In order to expand the set of words which were investigated, I have employed the synonym search function available on the Hansard corpus—which, usefully for these purposes, is a data set which has been semantically tagged. Whilst the tagging helps to get at synonyms and thus widen the lexical field of blame, it still has to be sanity-checked. To illustrate this issue, one of the words in blame lexical field from the Shipman Inquiry is failure. Searching for synonyms tagged in the Hansard Corpus to extend the lexical field generates: bankruptcy, insolvency, stoppage and flop—words particularly associated with financial failure. Whilst these may be found in the inquiry reports, they are unlikely to be associated with blame. Because of this, it is important that the suggestions which are generated by the corpus are not just used without due consideration.
Table 5.1

Frequency of blame-related lexical items, per million words

Token(s)

Inquiries

Hansard

Significance

Category A: Action of blame/criticism

Blame

30.7

46.8

$$\dagger $$

Censure

0.9

14.8

$$\dagger $$

Criticise

189.9

170.0

$$\dagger $$

Denounce

1.4

10.6

$$\dagger $$

Rebuke

0.6

3.0

$$\dagger $$

Reprimand

3.1

1.6

$$\dagger $$

Reprove

0.0

0.5

–

Total

226.6

247.3

$$\dagger $$

Category B: Attribution of responsibility

Answerable

3.2

4.9

$$\dagger $$

Culpable

2.6

1.8

–

Liable

26.6

93.2

$$\dagger $$

Responsible

600.0

414.2

$$\dagger $$

Total

632.4

514.1

$$\dagger $$

Category C: Framing of offence

Error

48.6

38.8

$$\dagger $$

Fail

389.7

198.0

$$\dagger $$

Fault

21.5

42.9

$$\dagger $$

Improper

5.6

15.9

$$\dagger $$

Inaction

7.9

2.1

$$\dagger $$

Inadequate

45.0

36.8

$$\dagger $$

Inappropriate

39.5

12.8

$$\dagger $$

Indecision

0.4

1.2

$$\dagger $$

Misjudg(e)ment

1.7

0.7

$$\dagger $$

Mistake

78.1

75.2

–

Offence

108.6

128.9

$$\dagger $$

Unsatisfactory

25.0

34.1

$$\dagger $$

Wrong

114.1

198.8

$$\dagger $$

Total

885.8

786.2

$$\dagger $$

$$\dagger $$ indicates that the difference between Inquiry reports and Hansard is statistically significant ($$p<0.01$$)

In widening the lexical field, and examining the frequency of these lexical items in the inquiry reports corpus, we should be able to gain a good insight into how often blame is either done, or at the very least, talked about at the conclusion of public inquiries. Table 5.1 shows the frequency of use of the blame lexical field in the public inquiry reports corpus, as compared to Hansard. The searches conducted to arrive at these frequencies (per million words) were lemma searches and so include more than just the token listed. For instance, criticise included criticise, criticising, criticised, as well as the related adjective critical. This allows us to see the broadest use of the lexical items under discussion. As I did with the lexical items associated with blame at the Shipman Inquiry, I have sought to group like lemmas together. As such, Table 5.1 shows items which frame the negative action, ones which involve the responsibility of the offender and others which convey an action of criticism or blame performed by the report’s author. This categorisation is not always straightforward and in context there will be occasions when the use of one lexical item will not necessarily be best described by the overall category, but the generalisations which obtain help in our understanding of what is going on in the inquiry reports. A log-likelihood measure of statistical significance was carried out on the raw frequencies, and a $$\dagger $$ indicates if there difference had a log-likelihood score of more than 6.63, meaning a p-value of less than 0.01. Emboldening of the frequency per million words shows which corpus has the most instances of a particular token (where the difference is statistically significant).

I first have to note some caveats to the figures which come out of the corpus searches. There are instances in which the inquiry chair is reporting on the view of some other party, and so the use of the lexical item cannot be said to be the inquiry performing the action in question. For instance, in the following example the lexical item critical reports on the feelings of a witness, rather than the attitude of the inquiry panel (emphasis added for clarity):
../images/375431_1_En_5_Chapter/375431_1_En_5_Figq_HTML.gif

Examples such as this serve to inflate the number of hits and may give the impression that the inquiry is engaged in more blame than it actually performs. The only way to exclude all of these instances of reported emotion or action would be to manually sift through every example—given that we are talking about hits of over 20,000 blame-related lexical items in the public inquiry reports corpus alone, this would be unfeasible. This is perhaps made still more clear when we consider that the Hansard corpus contains over 2.5 million such hits. This is also less of a problem than one might imagine. Since the comparison is made with the Hansard corpus which also contains hits which are reports rather than the personal use of the speaker, the comparison is a fair one. I am only talking in terms of comparison, either between the inquiry reports and Hansard, or between the categories within the blame lexical field.

A further problem with this approach is that sometimes the use of an item from the blame lexical field does not report either the chair’s view, or the view of a third party, but is used to discuss the idea abstractly, as in (94).
../images/375431_1_En_5_Chapter/375431_1_En_5_Figr_HTML.gif

Again these ‘meta’ uses, in this case, presenting a discussion of the problems potentially associated with blame, inflate the frequency of use of items in the blame lexical field. This is problematic if the numbers in Table 5.1 were used to explain how frequently blame was done in the public inquiry reports. But the data are used in a much more nuanced way. In what follows, I will talk about the frequency with which blame is talked about, as well as done in the inquiry reports. Later in this chapter, I will return to the meta-discussions associated with blame and what this can tell us about the preparedness of inquiry chairs to engage in blame.

With these caveats in mind, Table 5.1 shows a number of interesting differences, the three most important of which are:
  1. 1.

    Tokens which can explicitly show that the chair is engaged in criticism or blame are used less frequently in the inquiry reports than they are in the Hansard corpus;

     
  2. 2.

    Whilst fewer tokens for attributing responsibility (category B) were searched for, these are more frequent in the inquiry reports than are blame action tokens (category A); they are also more frequent in the inquiry reports than in the Hansard corpus;

     
  3. 3.

    Lexical items which are associated with spelling out the nature of the offence (category C) are most frequent (both in terms of number of items belonging to the lexical field, but also in terms of frequency). They appear significantly more in the inquiry reports corpus than in the Hansard corpus.

     
That items like blame, censure and rebuke should occur less frequently in the inquiry reports than they do in Hansard, and less frequently than the other categories in the blame lexical field speaks to the implicit nature of blame which I have been trying to draw out in this chapter. The authors of public inquiry reports do not frequently engage in explicit, incontrovertible, straightforward blaming. If we look at the uses of the lemma blame, of the 354 instances in the public inquiry corpus, only 39 are used performatively—i.e. to explicitly carry out the action of blaming, as in example 95:
../images/375431_1_En_5_Chapter/375431_1_En_5_Figs_HTML.gif

Examples such as (95) show who is to blame and for what. But what is noticeably absent—and would be seen in a more typical performative statement—is who is doing the blaming. I would still contend that an utterance like this is still performative (it ‘does’ blame), but it is left implicit that it is the view of the inquiry panel that takes the view that blame is necessary. The avoidance of using blame in a construction that requires a clear subject (i.e. a blamer) is found in a number of the 39 performative uses of blame. Again, this reflects a possible reluctance on the party of the inquiry panel to engage in blame. Framing blame in this way may also give the impression that such a judgement is almost an inevitability—it is a decision which has not simply been of the panel’s making, but (in this case) anybody in possession of the facts would also blame the Coroner for failing to do their job properly.

The fact that we see steadily more instances of tokens on the more implicit end of the blaming spectrum also points to the idea that inquiries prefer to avoid explicit blaming. There are more instances per million words of tokens related to responsibility for an action compared to blame for it, and still more examples which frame an action as bad. The latter requires more work on the part of the reader, who has to do some reading between the lines—not only to connect a potential actor to the negative action, but also to interrogate why it is relevant for the inquiry chair to report on this negative action.

In comparison to Hansard, we find that category A tokens are used significantly less frequently in the public inquiry reports. For the reasons stated earlier, I cannot say whether this discrepancy means that politicians are engaging more in blaming and criticising, but it does suggest that this is being talked about more frequently in this setting. This is surprising—given the focus of public inquiries is exclusively on wrongdoing, one might imagine this aspect of the lexical field would occur more frequently than in Parliament, where the range of areas for debate is so much greater. That we do find a higher use of the more implicit aspects of the blame lexical field in the public inquiries than in Hansard is further evidence for what I have been arguing. It is not that blame per se is avoided in public inquiry reports, it is explicit blaming which is eschewed. The reader of inquiry reports is left to ‘read between the lines’ about the fact that individuals are being blamed. I believe this stems from two factors: one which I have already noted and following Edwards and Potter (1992), blamers do not want to look like interested parties; the other which I will talk about shortly—blame can be presented as a negative reaction, and so to be engaged in explicit blame might be viewed as hypocritical.

5.5 Modality

In Sect. 5.3.7, I mentioned the use of the modal verbs should and ought to which were used by the inquiry chair to present an ‘alternative world’ in which the right thing was done by a witness to compare to the reality in which it was not (see Jeffries (2010: 150ff.) for a discussion of modality in relation to possible worlds (Ryan 1991) or text worlds (Werth 1999)).

In this section, I want to discuss other aspects of modality and its potential for blame. I will start by first outlining the different types of modality and views in previous literature on their relative strength. I will then go on to discuss the use of modality in the wider inquiry reports corpus and show how it gives further evidence for the implicit nature of blame in the reporting stage of a public inquiry.

Huddleston and Pullum (2002: 173ff.) offer an extremely clear account of modality. They argue that modality allows the speaker to express their ‘attitude towards the factuality or actualisation of the situation expressed by the rest of the clause’ (ibid.). Charteris-Black (2014: 113f.) suggests that there are two main types of modality epistemic and deontic. Epistemic modality is concerned with the likelihood with which an event described in the utterance is to occur (or has occurred). Deontic modality refers to obligation or necessity—how much a speaker or other party is obligated or required (or has been) to carry out the action described in the utterance. To these two categories, we can add a third type of modality—dynamic. Dynamic modality relates to the ability of a speaker or other party to carry out an action under discussion. Huddleston and Pullum (2002: 179) note that the boundary between deontic and dynamic modalities is fuzzy (and we will see this shortly).

Charteris-Black (2014) in his study of modality considers aspects outside of modal verbs and rightly includes these in his charting of high, medium and low modality. That is to say, certain expressions outside of modal verbs commit the speaker to varying degrees to the statement they are producing. For instance, ‘certainly’ suggests high commitment to the truth of an utterance, ‘probably’ conveys a medium degree and ‘possibly’ commits the speaker only to a low degree (Charteris-Black 2014: 113). I agree with Charteris-Black that modality can be conveyed by expressions outside of modal verbs, but the English language is so rich in these expressions that it makes a straightforward investigation of them difficult.

In what follows, I narrow my focus onto three modal verbs which allow speakers to express their attitude with varying degrees of strength towards the proposition contained in their utterance. I will discuss must (strong), should/ought to (medium) and may/might (weak). I will explore what verbs these modal expressions in past time collocate with and then discuss what this can tell us about blame. Note also that I will be exploring third person uses only, as these are the uses that would be associated with blame (rather than self-blame which we would find with first person uses).2

5.5.1 Must Have

I will start by looking at the strongest modal verb ‘must’, which in its epistemic use expresses that idea that something is surely or certainly the case. This epistemic use is much more common in past time than deontic uses expressing obligation or necessity. ‘Must have’ collocates most frequently with the verbs shown in Table 5.2.
Table 5.2

The frequency of the top 10 verbal collocates of ‘must have’, frequencies per million words with raw frequencies in brackets

Collocate of ‘must have’

Frequency pmw (raw)

been

46.8 (539)

known

3.4 (39)

had

3.1 (36)

come

2.7 (31)

seen

2.5 (29)

fired

2.1 (24)

taken

1.9 (22)

given

1.4 (16)

made

1.2 (14)

realised

1.1 (13)

But there must be further interrogation of the data which arises from these searches, and doing this shows the perils of an approach which neglects qualitative consideration of corpus data. In looking at the examples which include must have been, it is clear that many of these do not really give rise to an implicature of blaming. Consider, for instance, the following example from the Bloody Sunday Inquiry which is representative of many of the examples of must have been:
../images/375431_1_En_5_Chapter/375431_1_En_5_Figt_HTML.gif

Here is is clear that the idea which the chair is trying to convey is that ‘it is doubtless the case that X’. However, because this is example is in the passive voice, the subject most of the people does not refer to the agent of the action (i.e. the people who deliberately performed the action). As such the people who were moved away are not viable blamees. Because of this, such an example does not give rise to an implicature of blame.

Contrast that example with the following which also contains must have been. This example comes from the Baha Mousa Inquiry which investigated the treatment of prisoners during the Iraq War and was named after a prisoner who died whilst in the custody of British soldiers in Basra, Iraq. In this example, we do not find the passive voice:
../images/375431_1_En_5_Chapter/375431_1_En_5_Figu_HTML.gif

In this case, the grammatical subject and the semantic agents (or possibly more accurately, the experiencers, i.e. the people who have a cognitive experience) are one and the same. The same meaning of ‘it is doubtless the case that X’ is conveyed by must have been here, but because of the experiencer is a potential blamee, it also generates an implicature that the conduct of the subject is blameable.

This is perhaps even more clear with some of the other verbs which must have collocates with frequently. The next most frequent verb is known. The difficulty relating to active vs. passive does not arise in these cases. The following example is representative of the 39 involving known:
../images/375431_1_En_5_Chapter/375431_1_En_5_Figv_HTML.gif

Once again, the modal constructions can be taken to mean something like ‘it is inconceivable that Moutarde did not know of the conditions...’. In this case, the chair is either directly contradicting the impression given by a witness, or is coming to a view based on the weight of all of the evidence provided to the inquiry. Either way the modal closes off the potential blame avoidance strategy of suggesting a lack of knowledge of the negative. In this example we see the modal must supplemented later with ‘ought to’ which gives the impression of what a right-thinking would have done with this knowledge. From this the reader can derive the view that Moutarde is blameable.

Overall, excluding ‘must have been’ and including verbs outside of the top 10 collocates, must have + verb appears approximately 49 times per million words. This is more than all of the category A blame-related lexical items except criticise outlined in Table 5.1.

5.5.2 Should/Ought to Have

Moving on to medium strength modal verbs—should and ought to are largely synonymous and I will treat them as such here. Unlike must, the deontic use of these modals has primacy especially since ‘[a]n epistemic reading is hardly possible with past time situations’ (Huddleston and Pullum 2002: 186) such as those we are interested in in this section. As such, the examples which arise with these modals relate to obligation or necessity, rather than probability or possibility.

Let us first look at the frequency of use of these modal expressions in the public inquiry reports corpus. Having explained the difficulties which arise with modal expressions in the passive, I have excluded ‘been’ from the results shown in Table 5.3,3 which otherwise shows the first ten verbal collocates of the medium strength modal should/ought to.
Table 5.3

The frequency of the top 10 verbal collocates of ‘should/ought to have’, frequencies per million words with raw frequencies in brackets

Collocate of ‘should/ought to have’

Frequency pmw (raw)

done

18.5 (214)

taken

9.3 (107)

made

8.0 (92)

realised

4.6 (53)

given

4.3 (50)

referred

3.9 (45)

had

3.6 (42)

reported

3.2 (37)

ensured

3.2 (37)

asked

3.2 (37)

But looking just at the top 10 verbs masks the fact that ‘should/ought to have’ collocates with many more verbs less frequently (i.e. it has a higher type/token ratio). In total these modals used in past time in conjunction with verbs except ‘been’ occur 296.7 times per million words. This is nearly six times more frequently than the stronger modal ‘must have’. In looking at an example of the use of ‘should have’ I will seek to show why it tends to be favoured. The following example comes from the BSE Inquiry which set out to find the causes behind the spread of Bovine Spongiform Encephalopathy (more commonly known as Mad Cows’ Disease).
../images/375431_1_En_5_Chapter/375431_1_En_5_Figw_HTML.gif
This deontic use allows the chair to state, almost a general rule, what would be considered the right thing to do in a given situation. In this case, the general rule that is conveyed by the chair is that delays in communicating the risks associated with potential biohazards are to be avoided. Moreover, the example also demonstrates that:

[w]ith past or present time [deontic uses of should/ought to] are commonly used when it is known that the situation was/is not actualised, in which case they convey criticism. (Huddleston and Pullum 2002: 186)

So, whilst should/ought to express medium modality, the forms found in the inquiry reports are intrinsically associated with criticism, albeit that this still remains implicit rather than explicit. The use of a stronger form of modality would not convey that the action in question failed to be realised, and thus would potentially not necessarily convey the criticism which the chair is seeking to provide. As such, the notion of strength of modality is less relevant when we think about the association of modal verbs with blaming.

5.5.3 May Have

The final modal verb which I will look at in this section is may—the weakest of the modals looked at here. For the most part, the use of may in the past time generates an epistemic reading and one which does not strongly commit the speaker to the proposition produced.

Table 5.4 shows the ten most frequent verbs which collocate with may have:
Table 5.4

The frequency of the top 10 verbal collocates of ‘may have’, frequencies per million words with raw frequencies in brackets

Collocate of ‘may have’

Frequency pmw (raw)

had

8.5 (98)

contributed

3.1 (36)

caused

2.9 (33)

made

2.4 (28)

infected

2.4 (28)

seen

2.2 (25)

occurred

1.9 (22)

taken

1.8 (21)

resulted

1.7 (20)

led

1.6 (19)

If we consider all verbs, even those which give just one hit, then ‘may have’ is used 106.7 times per million words—making it more frequent than the strong modality of must have, but less frequent than should/ought to have.

In thinking about why that might be, the following example gives us a good idea. Example 100 comes from the Victoria Climbié Inquiry, which I will explore in a little more detail in the next chapter. The inquiry investigated the role of public services in failing to protect a child who was tortured and murdered by her guardians. The following example discusses the role of social services in doing background checks which could have flagged concerns about Climbié’s guardians:
../images/375431_1_En_5_Chapter/375431_1_En_5_Figx_HTML.gif

Example 100 shows that the use of ‘may have’ does not strongly commit the chair to the idea that Mr. Victor having acted differently would have resulted in a different outcome. This is especially the case because of the complex chain of events which behaving differently would have triggered—each stage possibly opening up an opportunity for more information to have been uncovered. But the chair does not explicitly state that even if more information had come to light that this would have had a material impact in the case. As such, it really is left up to the reader to come to a conclusion about whether the conduct of Mr Victor is blameable, and they may do this on the basis of the flowchart which I presented in the previous chapter in Fig. 4.​2. This ‘passing the buck’ to the reader when it comes to deciding on blame may account for why may have occurs more frequently than must have—it allows the inquiry chair to suggest blame, but plausibly deny that that was the intention if that were necessary. The reasons for this desire to deny blaming will be drawn out in the next section.

5.5.4 Summary

This section has shown the important role that modality plays in the discursive construction of blame. Modal verbs allow the chair to express the credibility of the evidence which came before the inquiry, the knowledge of witnesses with respect to blameable conduct and what a ‘reasonable person’s’ conduct is likely to have been when presented with similar situations. The contrast between reality and this imagined, more positive, world is made clear when using modal verbs (although I am sure other expressions of modality would also contribute to this discursive construction). However, it is left to the reader of an inquiry report to infer that the intention of the chair in drawing out this contrast is to blame a party or parties invoked. The strength of modality going from strong (must) to weak (may) either affects the likelihood with which a reader infers blame, or where on the ‘blame potential’ continuum the reader places the blamee. That having been said, the relationship between strength of modality and blame is not a straightforward one, as we have seen. Stronger forms of modality may not necessarily mean stronger forms of blame—should/ought to is particularly associated with criticism and so this may affect a reader’s judgements around blame potential. Once again, we have seen a preference for inquiries to prefer implicit blame over explicit. In the next section I will explore why this might be the case.

5.6 Views of Blame and the Counter-View

When exploring the Shipman Inquiry, we found that the inquiry chair took the responsibility of blaming seriously. She recognised that finding parties responsible for actions which, had they not been committed, may have prevented the murders of three of Shipman’s patients, was condemning them to a life in which they had to carry this huge burden. Whilst sympathising with them, she re-iterated that this blame was of their own making. This was not the only inquiry in which a view of blame was presented by the inquiry chair. Indeed, I noted earlier that the uses of the word ‘blame’ that we see at inquiries are mostly (though not wholly) used non-performatively—i.e. blame is not used to ‘do’ blame. Instead, blame is found most frequently in discussions about ‘blame’. In this section, I will take some time to unpick a view presented of the notion of blame and what is can tell us about why the action of blame is most frequently done implicitly, as I have just shown. I also want to spend some time exploring why I think some criticisms of blame are faulty and offer a counter-view that blame can be healthy.

The following lengthy extract from the report of the Mid-Staffordshire NHS Foundation Trust Inquiry, which sought to investigate poor standards of care in a group of hospitals, summarises discussions about blame found in a number of inquiry reports (including the Bristol Royal Infimary Inquiry and the Victoria Climbié Inquiry) .
../images/375431_1_En_5_Chapter/375431_1_En_5_Figy_HTML.gif

This extract from the Mid-Staffs Inquiry report firstly draws attention to the idea that wanting to blame is a normal human reaction following negative events, but goes on to argue that this is something which should be resisted by a public inquiry. This, clearly, is not something which other inquiry chairs agree with—we saw earlier in this chapter an explicit section of an inquiry report dealing with blame, and we have seen that other inquiries also engage in blaming. In resisting assigning blame, the inquiry chair acknowledges that this goes against the desires of some who pressured for the inquiry to be established initially. Whilst the motivation for this might be a reasonable one—that assigning blame may cause further harm—I would suggest that failing to fulfil the desires of those who called for the inquiry is not without risk. Indeed, failing to meet the expectations of those who have been wronged risks giving rise to a feeling of being wronged a second time.

Various reasons are set out for avoiding assigning blame which I will take in turn. The first is that in the particular case looked at by the Mid-Staffs Inquiry there were so many instances of blameable conduct that it becomes ‘futile’ to engage in blame, instead the system in which those individuals were operating can be said to be blameworthy. I find this difficult to agree with—to take such a statement to its conclusion, it would mean that blame is avoidable if enough people are involved in wrongdoing. That means that people are more like to ‘get away with it’ if the scale of wrongdoing is larger. That seems an untenable situation.

This does not deny the very fair point which the chair makes here; system failures were behind much of this wrongdoing. But we have seen that this can stand in mitigation for blame—it can provide an explanation for the negative actions (whether intention or unintentional) that people were engaged in. This may reduce their blame potential. But I do not think it has to short-circuit blame in the way that the chair suggests.

The chair also presents what he considers to be the risks associated with blame. Firstly, that emphasising the blameworthiness of individuals can give a faulty impression that all that needs to change with the (broken) system is for those individuals to be removed. Secondly, if people see that they are likely to be blamed then it makes them more likely to engage in covering up innocent mistakes rather than being open and honest about those mistakes and treating them as a learning opportunity. These are certainly risks with blame and they have been identified in various places, particularly in the field of medicine (see amongst others: Bond 2008; Catino 2009; Pearn et al. 1998). However, there are risks in not blaming—aside from a feeling that justice has not been done amongst people who have been wronged. If there are no clear mechanisms for holding people to account (which the public inquiry has been argued to be, as discussed in Chapters 1 and 2), then this risks potential wrongdoers having a sense of impunity. The potential negative outcomes of this are obvious—it was a key finding in the Royal Liverpool Children’s Inquiry and the Shipman Inquiry —that the doctors implicated in those inquiries were able to act unchecked and with a sense that they would not be caught. This resulted in further wrongdoing. As Hood (2011: 185) notes:

where complex organizational or institutional arrangements are concerned, agency will almost always be ambiguous too. Short of some imagined nirvana in which those difficulties could be made to disappear by a technical fix or a general outbreak of social altruism, how are societies to decide what constitutes an avoidable loss and who or what is responsible for such a loss?

In other words, there has to be an accountability mechanism, even if that is difficult when faced with large organisations.

But the views presented by those who see blame as potentially negative do go some way to explaining why we see explicit blame so infrequently in inquiry reports. It may be seen as hypocritical on the one hand to talk about the risks of blame, only to then go on to do this elsewhere in the inquiry report. I argue that this risk of hypocrisy is why implicit blame, using modality, amongst other things, to draw attention to blameworthy conduct, is favoured.

5.7 Conclusion

In this chapter, I have demonstrated that explicit blame is rare. Even where an inquiry included a section on who was to blame, the performative use of the verb blame was absent—instead the report relied on the textual coherence of the title ‘Who was to blame?’ and the discussion which followed to ‘do’ blame. There are a handful of examples of performative blame in the full inquiry reports corpus, but these are marginal. Blame is instead used to discuss the risks associated with blaming, the understandable desires of some to assign blame and provide explanations for why blame is not possible.

In using the explicit blaming section of the Shipman Inquiry report, I was able to develop a lexical field of blame—words which are inherently associated with blaming, and capable of performing this action implicitly. But we found that even these words associated with blaming are underused in the inquiry reports, in comparison to another corpus. In part, this is explained by the fact that some inquiries eschew giving blame—the chair believing that blaming can cause problems in the future, creating a culture in which mistakes cannot be openly discussed for fear of being condemned. The argument put forward by some is that the learning of lessons—what I mentioned as being a key focus of some inquiries in Chapter 2—is made more difficult if a chair blames inquiry participants.

I challenged this, and argued that blame is important—it is an accountability measure; that there is an obvious consequence to wrongdoing may help to prevent future wrongdoing. In other words, the personal accountability that comes from the risk of blame may mean that someone is more motivated to learn the lessons which a previous inquiry has sought to make obvious. Without a personal risk, a future wrongdoer may feel able to act with impunity. This is not a good basis for civic society to operate under. Moreover, without blame, those who have been wronged may feel ‘short changed’ by the justice system. Since the public inquiry is often the end point in a campaign for truth and justice, a failure to blame those responsible for hurt and suffering may result in disenchantment with the process. This should also be avoided.

It is for this reason that I believe the Shipman Inquiry is a good model for others to follow. Dame Janet Smith made it clear who was blameable and being blamed. But simultaneously, she did so in a way which acknowledged there was some mitigation and which made clear her understanding of the implications of her blame for those being blamed (as well as those affected by the wrongdoing). This directness—albeit done with sensitivity—is better, in my view, than what we do find in all of the inquiry reports. I showed that modality, especially speaking in terms of what should have happened, allows the reader to infer that the wrongdoer is being castigated for not having done this. The use of other modal markers like must and may are also indicative that blame is incipient or embedded within the description of events. The use of the passive with modal verbs means we cannot always infer blame, however. The reasons for chairs going about blame in this way include legal considerations: criminal trials may be prejudiced by explicit blame—this likely explains why the Robert Hamill Inquiry has yet to publish its report. But that does not explain the non-assigning of blame for all inquiries, since they do not all involve criminal behaviour and still others involve potential blamees who are dead.

What is needed, I suggest, is further work to explore whether these legal concerns around prejudice can be surmounted. This is naturally beyond the scope of this linguistic work. It may also be that amnesties from prosecution, such as those offered by the South African Truth and Reconciliation Commission (see Tutu 1999), would be considered preferable by some victims and family members. But this would require the government to develop a new Inquiries Act, one which allowed for the involvement of those affected by the events investigated by an inquiry in its establishment, procedures and Terms of Reference. I submit that this would be a step in the right direction, if the public inquiry is to continue to have value.

In the next chapter, I explore a final aspect of public inquiries and particularly focus on what happens on their completion: apologies for the wrongdoing uncovered. I will discuss what the speech act of apologising has to do with the discursive construction of blame, and what an apology’s value is for civic society.