Two cases were posted for consideration. The first was:>(1) A person faces
trial for a public order offence where the definitionof the offence includes an
element that concerns the likely impact of theirbehaviour on bystanders (e.g.
whether a bystander would be threatened, alarmed etc.).Two of the bystanders to
that alleged public order offence have importantengagements abroad on the date
of the person’s trial and do not want to attendand the person applies for
witness orders for those bystanders to attend andgive evidence for the defence
(on the basis that the bystanders may say theywere not threatened, alarmed
etc.).> Call this 1A. It may help understand the problem in 1A to widen out
theproblem. Here is another case. Call this 1B.
1B. An applicant has a discrimination case arising out of an allegedclandestine
difference in the rate of pay between themselves and a comparator.The employers
say the comparator got some extra pay for working extra hours butmaintain there
was no difference in respective rates of pay for hours actuallyworked. Accounts
prepared by the employer’s bookkeeper show both applicant andcomparator being
paid at the same rate for hours actually worked. These sumsare free of
mathematical error. The same accounts also detail the payments forthe
comparator’s supposed ‘extra hours’: these sums contain ‘errors’, in thatthe
comparator is paid amount x forthese extra hours when he should have been paid
amount y given the rate of pay. Not only is x incorrect as the sum due for the
extra hours but x is exactly the extra amount that thecomparator would receive
if they were working no extra hours but were beingpaid at a higher percentage
rate than the applicant for their standard hours. So the applicant’s case in
respect of the accounts is that they have beendrawn up to disguise the fact the
comparator is being paid at a higher rate forhours actually worked – disguised
by paying the difference (between the higherand lower rate) in the form of a
sum for ‘extra hours’ (that have not actuallybeen worked). The employers don’t
propose to call the bookkeeper as a witness. Theapplicant seeks a witness order
to call the bookkeeper with the intention ofquestioning them on the ‘errors’ to
show the accounts contain deliberatefabrications (unless the bookkeeper admits
the deliberate fabrications, thismay involve having the bookkeeper declared
‘hostile’ so the applicant caneffectively ‘cross-examine’ their own witness).
Should the witness order be granted in 1A or 1B (both, neither, only one)?
We are asking this question in the context of evaluating whether it isbetter to
understand the law as an exercise in ‘conceptual analysis’ or
in‘problem-solving’. So we can further ask: how far does CA takes us in
answeringthis question?
Not far. Not any distance really. If it did we should be able, by CA, to answer
the question at thispoint – or, if not at this point, we should be able, by CA,
to arrive at an answer somewhere along the line below. Mysuggestion is that (a)
there is no point along the line below that there is a eurekamoment where the
answer arrives by CA and (b) all the along the line below weare developing our
understanding in terms of ‘problem-situations’. What we findis that the form of
the problem-situation in 1A and 1B differ when we getbeyond the conceptual bare
bones to the specifics, and that these specifics (qua ‘problem-situation’) may
play adecisive part in what rule is adopted. So even though the rule adopted
mayapply to both 1A and 1B, it may very well be that whatever rule is adopted
isaffected by whether the problem presents itself in the form of 1A or 1B.
Thiskind of contingency cannot easily be accounted for in terms of CA though it
caneasily enough be explained in terms of a problem-solving approach.
So, when considering the following, bear in mind that there is a
constantinvitation to defenders of ‘law as CA’ to jump in at any point and
explain“Well, from this point on we can of course arrive at the answer by CA as
may beshown by…..” My claim is that this kind of explanation cannot plausibly
begiven and that trying to recast the whole story in terms of CA is ‘post hoc
pseudo-explanation.’
1A and 1B, two seemingly disparate cases, become related because the
ruleadopted in one of them affects the other. The order and hierarchy in which
theyare presented before the courts may have a significant effect on which rule
isadopted at a high level within the legal system (for CA no more explains
legaldevelopment than it explains the ‘order of play’ in the ‘War of the Roses’
andits effect on outcomes). As it happens, 1A was presented first and gives us
arule that it is still showing its teeth some 20 years later e.g. in the
jurisdiction of Hong Kong it was applied to dismiss an appeal as recentlyas
2010 [HKSAR v Au Kwok Kuen andothers]. A 1A-type case has led a certain rule
becoming embedded. Itmight all have been very different – but CA cannot explain
this as a ‘problem-solving’approach can (see below).
Before looking further at 1A, we might note how different the law might
havebeen had 1B presented itself first (say, on an appeal against refusal of
awitness order in a 1B situation). For had 1B become the subject of a
decisionof the Court of Appeal (or even Supreme Court), it may well have been
that therule adopted would have been that the witness order should be granted
on thefollowing basis: though the bookkeeper may prove ‘hostile’ or
unfavourable tothe applicant’s case in terms of whatthey say absent
‘cross-examination’, their evidence subject tocross-examination may be
discredited in a way that supports the applicant’scase (perhaps even
decisively). In other words, in a 1B-type case the court mightwell have viewed
it as manifestly in the interests of justice that an applicantbe allowed to
compel an unfavourable or hostile witness in order to discredit their evidence
in a way that lends support to theapplicant’s case. Allowing a ‘hostile’
bookkeeper to be cross-examinedmight advance the applicant’s case significantly
so that instead of the courtfinding "There were ‘errors’ but the bookkeeper was
not called and we don’tknow what explanation they might have given" the court
would find "There were ‘errors’and the bookkeeper was called but could give no
credible explanation to counterthe suggestion the ‘errors’ were part of
deliberate manipulation of theaccounts."
There is very different flavour to the 1B case than to 1A when we flesh outthe
facts (see below). 1B ‘presents’ as a case where a witness order is necessary
to stopthe other side ‘hiding their witness’, by not calling them, when that
witness wouldfall apart on cross-examination. In terms of the interests of
justice, the desirability of the witness order is reinforced where the witness
is (or may be) partyto fraud or other discreditable conduct, including what
would be a fraud on thecourt or perverting the course of justice. By contrast,
as shown below, 1A ‘presents’as a case where allowing a witness order would
allow the applicant to make aneven greater nuisance of themselves than they
have already and to use theprocess of the court to pester or disturb by calling
someone as a witness whenthis could not realistically help their case, and also
allow a witness order tobe granted for a possibly “ulterior purpose”. In other
words, in 1A there is nogenuine ‘interest of justice’ served by calling the
witness and calling thewitness borders on ‘abuse of process’.
It is 1A not 1B that made it to the higher courts. In 1A thecourt did not view
it as manifestly in the interests of justice that the applicantbe allowed to
call an unfavourable witness: Rv Marylebone Magistrates’ Court, ex parte
Gatting and Emburey (1990) 154 J.P. 54. The application forwitness orders was
denied. The court ruled that a party cannot obtain a witnessorder where the
witness is not likely to be favourable to that party in theirevidence. Given
the rule in Gatting is sufficiently broad to apply in cases of both type 1A and
1B, we might note its central weakness asa rule: it overlooks cases, like 1B,
where a hostile or unfavourable witnessmay nevertheless be discredited in a way
that supports the case to which theyare hostile or unfavourable, and it
overlooks other cases where the hostile witness might support that case in
part. It is very possible this is because in its ‘characterisationof the
problem’, the court in Gattingdid not consider cases like 1B but simply had in
mind cases of the 1A type. Itis also possible this narrow characterisation was
not fully thought out becausethe court was fixated on the possible ‘abuse of
process’ in cases of the 1Atype.
Gatting and Emburey may not be well-known names in America and arelargely (and
perhaps best) forgotten here. They were cricketers who played for England.
Theyalso went on one of the ‘rebel tours’ of apartheid South Africa. In advance
of this, someprotesters mounted a pitch invasion to protest Gatting’s and
Emburey’s proposed‘rebel tour’ and found themselves charged with public order
offences. Theprotesters’ trial ended up scheduled for the period during which
the ‘rebeltour’ took place: if the witness orders were granted, the protesters
wouldsabotage Gatting’s and Emburey’s participation in the ‘rebel tour’. Howzat.
The problem for the court was that, though it noted its “suspicionin this case
that the process of this court is being abused,” the applicationfor the witness
orders fell short of a clear-enough ‘abuse of process’ (I wasin court for this
hearing and the judges certainly grilled the lawyers asregards ‘abuse of
process’; but, fundamentally, the applicants were notresponsible for the
scheduling of their trial, and there was insufficient proofthat the witness
orders were sought not for “the simple and proper purpose ofobtaining evidence,
but for a different and ulterior purpose”.) Unable to find ‘abuseof process’,
the court sought another way to refuse the application for thewitness orders –
by adding to the established condition that a witness hasrelevant evidence by
way of a further condition that the evidence be likely helpful or supportive to
the partyseeking the witness order. The court then found it safe to conclude
that neither Gatting nor Emburey were likely to be helpful.
In the immediate term, problem solved. And the Gatting rule can be defended
asin ‘the interests of justice’ for cases of the 1A type – where there is
norealistic prospect that the witness’ evidence would help the case of the
partyseeking the order. The problem is that the Gattingrule also applies to
cases of type 1B, and here the rule couldwork very much against the ‘the
interests of justice’. This creates an ongoing ‘problem-situation’ in W3.3
termsthat may yet give rise to an instance of a 1B type case that leads to an
appealfor reversal or amendment of the Gattingrule. ‘The interests of justice’
in both kinds of case might be better servedby an ‘amended Gatting rule’ viz.
that a witness order should not begranted where there is insufficient
likelihood that the evidence will assistthe party seeking the order but an
order may be granted in respect of a hostileor unfavourable witness where the
discrediting of their evidence may assist thecase of the party seeking the
order or where their evidence, though ‘hostile’in part, may have sufficient
parts that assist the case of the party seekingthe order. In outlining and
explaining above, we have followed a ‘problem-solving’approach – one that
allows that there may be different levels andaspects of a problem and that the
proper ‘characterisation of the problem’ isitself a central problem. My
suggestion is that this kind of account cannotplausibly be replaced by one in
terms of ‘conceptual analysis’. But let’s see. In the meantime perhaps we
should no more take seriously theclaim that legal argument owes to ‘conceptual
analysis’ (a claim derived fromthe continuing influence of ancient
“essentialism”, medieval scholasticism and themore recent ‘linguistic turn’)
than we would take seriously the piousdeclarations of lawyers centuries agothat
their legal arguments were derived from God.
DL
