In May 2008 barrister Mark Saunders was tragically killed by police after he repeatedly fired a shotgun out of the window of his Chelsea flat. During the subsequent Independent Police Complaints Commission (IPCC) investigation the officers involved conferred with each other in recording their accounts. Notwithstanding that the practice was still permitted by the Association of Chief Police Officers (ACPO), judicial review was sought on the grounds that in permitting joint note compilation the Commission had breached the European Convention on Human Rights, Article 2 for the protection of life, by making an effective investigation into the death impossible; the very conferring process, it was argued, must have contaminated, and therefore stultified any realistic scrutiny of, each officer’s account.

Risks of conferring prohibition

The Administrative Court rejected this on the basis that the proposed prohibition on conferring would have been likely to induce non-cooperation by the officers and would therefore have led to the investigation being less effective than that which was actually conducted: R (on the application of Saunders) v IPCC [2008] EWHC 2372 (Admin). This was later apparently latched on to by a “senior source” who thought that any ban on conferring would induce lawyers to tell officers not to say anything at an early stage (see the Guardian, 24/10/08).

At the same time, though, Underhill J, giving judgment, observed (at para 13): “However much an officer who has conferred with colleagues may strive to record only what he has seen or heard for himself, there is a real risk that his recollection will have been ‘contaminated’ by what he has been told; and he may in perfect good faith incorporate elements in his own account which have in fact derived from other witnesses, or subconsciously suppress elements which seem to him inconsistent with their accounts...There is also the risk that, quite apart from such innocent contamination, officers collaborating in producing their notes or statements may be tempted deliberately to produce an account which does not accurately reflect the individual recollections of each...[A]lthough the distinction...between innocent contamination and deliberate collusion is conceptually clear, its application may of course be a lot less clear in particular cases.”

Firearms Manual

The IPCC had for some time been encouraging ACPO to amend the conferring provisions in the Firearms Manual, but had recognised that this might have potential consequences for evidence gathering in other contexts. In the wake of the judgment ACPO has amended the Firearms Manual to state that in general there is no need for officers to confer with each other before writing up their notes about what happened and what was in their minds at the time force was used. However, it goes on to state that where conferring is necessary officers must, in order to ensure transparency and maintain public confidence, record the fact that some discussion has taken place, highlighting the time, date and place of conferring, the issues discussed and with whom, and the reasons for such discussion. This has not been welcomed by Paul Davis, Police Federation representative for firearms officers, who believes that conferring is “the best way of protecting members’ interests and getting to the truth”. He sees the amendment as “an open attack on the professionalism of officers”, tying their hands and leaving them vulnerable (the Guardian, 24/10/08). But what of transparency and the protection of the public?

History—law and psychology

Police officers customarily rely on R v Bass, [1953] 1 QB 681, in support of what they jealously cling on to as an inalienable right. Indeed, Underhill J (at para 59) noted that the “right to confer” in the context of police shooting inquiries had become “something of a talisman”.

In fact, Bass involved ex post facto notes of an interview with a suspect (now compulsorily tape-recorded under PACE), rather than an observed incident. More recently, pre-trial conferring, at least by civilian witnesses, has been judicially deprecated: see, eg, R v Richardson [1971] 2 QB 485 at p 490 (cautioning against mutual comparison by witnesses of their statements before giving evidence); R v Arif, the Times, June 17 1993 (forbidding witness rehearsal); R v Skinner (1994) 99 Cr App R 212 at p 216 (warning that police evidence based on the common practice of joint note compilation may be attacked as distorted by discussion); R (Green) v Police Complaints Authority [2004] 1 WLR 725 at p.747, HL (risk of “trimming” if witnesses see each other’s statements).
Most notably in R v Momodu and Limani [2005] 1 WLR 3442, at p 3453, para 61, Judge LJ, as he then was, referred to “the well known principle that discussions between witnesses should not take place and proofs of one witness should not be disclosed to any other witness” and he continued: “The witness should give his...own evidence, so far as practicable uninfluenced by what anyone else has said...the rule reduces, indeed hopefully avoids, any possibility that one witness may tailor his evidence in the light of what anyone else said, and equally, avoids any unfounded perception that he may have done...Recollections change. Memories are contaminated. Witnesses may bring their respective accounts into what they believe to be better alignment with others. They may be encouraged to do so, consciously or unconsciously. They may collude deliberately...Whether deliberately or inadvertently, the evidence may no longer be their own.”

The benefits of conferring

Underhill J (para 16) pointed out that “the advantages of a blanket prohibition on collaboration in the production of notes might very well involve disproportionate disadvantages” and a ban “not only would be difficult to enforce in practice but would in many cases have serious operational disadvantages [as where the] prompt exchange of information between officers in the immediate aftermath of an incident is often essential [although] that objection might not apply to collaboration in the production of notes”.

To be more specific, we would add that as a matter of everyday experience the very obvious benefit of conferring, perhaps its chief justification, is that the memory of a certain detail, initially not recalled at all, can be instantly evoked by the bare cue of hearing someone else mention it. Routine conferring by the police in compiling their notes has presumably been tolerated as an exception to the general ban because, unlike civilians—who will usually be recalling events which for them are unique—officers are witness to so many similar incidents that their apparent memory of an event may derive from the generality of their experience rather than the particular.
Given that for the time being police officers are likely to continue conferring over their evidence in criminal cases, it will be necessary to formulate: (i) suitable procedures for optimising the efficiency of the practice in retrieving genuine memory and for avoiding either innocent cross-contamination or collusive purported adoption; and (ii) appropriate forensic sanctions to attain compliance.


If conferring is to continue in all criminal cases its efficacy necessitates a two-stage process. First, witnesses must be required initially to record their account (either in writing or onto a sound recording machine) prior to any conferring with a co-witness. This will preserve the unadulterated or “pure” state of their memory before the influence of anyone else’s memory is brought to bear on their minds, “so that in cases where a real dispute emerges as to the precise sequence of events the extent and significance of any subsequent revisions in the witness’s account can be assessed” (Saunders, para 15). Second, the court must be afforded an unassailable means of scrutinising the actual conferring process in order that the exact course of the process of mutual memory cueing is comprehensively transparent. Self-evidently this must involve, at the very least, audio-recording of the session of “pooling recollections”. The process must be formalised: canteen conferring should be history.

“Safe” conferring—safe for whom?

In the absence of video- or audio-recording is there anything which can be done to refine the existing conferring process? What of a compromise—a ban on oral conferring but allowing officers, once they had completed their initial accounts, to read those of their colleagues? If this stimulated or further refreshed their memory, they would then write out a supplementary account explaining what had caused their earlier expressed memory to change or to be enlarged —and how. Such an approach is unlikely to be more transparent (and anyway would offend the general rule against mutual inspection of witness statements). An officer who reads a colleague’s note might easily be tempted to adopt an assertion missing from his own initial account if he appreciated its importance to the prosecution case or naïvely supposed that an exact identity of accounts would be more convincing than mild differences (rather than less). Such an officer could, with impunity, feign a spurious reason for having had his memory jogged. With conferring, the defence can at least cross-examine the conferring witnesses on their methodology. Better to leave things as they are if we can’t have an audio record.

Forensic strategies—PACE s 78

With the exception of R v Arif (cited above) none of the reported cases has involved an argument that evidence recorded during or following a conferring session should have been excluded from evidence in accordance with s 78 of PACE, on the grounds that the logical consequence of an undisclosed conferring session is that the evidence so obtained is incapable of reasonable or realistic scrutiny. If preliminary lone-composed accounts and audio recording of subsequent conferring sessions became standard practice, s 78 applications would become routine in non-compliance cases. Even if exclusion were unwarranted, non-compliant conferring could still be a matter for comment to the jury: see R v Arif.

Good sense, not panic

There can be no logical justification for treating police officers in such a radically different way from other classes of witness. By recognising the risks to justice involved in conferring, Underhill J and ACPO  may have opened the way to a more sensible and transparent method for police officers to prepare their evidential accounts. Now is the moment for police authorities, prosecutors and the courts to take their cue and introduce good sense and reliable methods into this highly controversial and much neglected area. There is no need to panic—the solutions to the problems created by conferring are not as elusive as might be feared. 

Anthony Heaton-Armstrong and David Wolchover are practising barristers, at 9–12 Bell Yard and 7 Bell Yard respectively. One of their earlier articles on this topic, “Knocking Heads Together”, appeared in Counsel in April 1993