A rock and a hard place: autism as behaviour v autistic conduct

Further thoughts on the lessons of W v H [2022] EWFC 150

As someone who lives closely with autism, I have a special interest in the way autism is understood within wider society. That understanding has come a long way in recent years. It’s commendable that judges are trained in the understanding and treatment of autistic people, as described by the judge in W v H [2022] EWFC 150, HHJ Greensmith, who has himself been involved in designing and delivering training on behalf of the Judicial College.

The first thing to understand about autism – as the judge references in this judgment – is that, when you meet one person with autism, you’ve met one person with autism (in the words of Dr Stephen Shore). That is to say, there is little to be gained from making generalisations; each autistic person sees the world in their own individual way and consequently has their own very specific needs if life is to be as manageable and as fulfilling as possible. The important thing is to understand what those needs are for the person in front of you.

Of course, the range of autistic traits is vast and might fluctuate – particularly under stress. We might all remember the kid at school who was a bit out of kilter, a bit isolated (maybe we were that kid), but now, looking back, we can see that that kid was probably autistic. Maybe he played alone, maybe she never quite seemed to “get it”. These people grew up, and in many cases, like the husband in this case, achieved success, and maybe managed to hide their quirkier behaviours from friends and colleagues. At the other end of the spectrum, there are those people who are never able to be independent, those who are non-verbal, and those whose behaviour is difficult to manage.

As Maisie Lockyer sets out in the previous article, the husband in this case was a doctor. He probably was one of “those” kids in school. He certainly didn’t know that he was autistic – not until long after the parties had separated, and some time after the wife had issued her petition for divorce, when he received an official diagnosis from his consultant clinical psychologist of level 1 autistic spectrum disorder (which might be described as the “mildest” form of autism). There is no detail in the judgment about the ways in which autism manifested in the husband, but at the ground rules hearing, which took place a month or so before the final hearing, it was determined that the husband needed:

  • An intermediary, who was invited to make suggestions as to how the husband might be encouraged to expand upon any monosyllabic answers to questions. (An intermediary is a professional appointed to support the witness in communicating effectively with the court; in this case, their costs were met by HMCTS.)
  • A McKenzie Friend.
  • To be cross-examined via questions which had already been put to him in writing and approved by the judge.
  • The option to attend the hearing remotely, with the intermediary at his side.

In considering the husband’s individual needs, the judge was bound by rule 3A of the Family Procedure Rules 2010 to consider how the husband (and the wife) were able to participate, and in this he was able to rely not only upon his own training and experience, but also on The Advocates Gateway, and, in particular, Toolkit 15: “Witnesses and defendants with autism: memory and sensory issues”.

It seems clear from the participation directions that the judge considered that the husband would struggle to answer (and possibly to understand) questions, particularly if those questions were asked in the usual way (without prior warning), and that in order to determine the issues fairly, the court would need the husband to be supported in the ways set out.

It’s really important, when dealing with clients who may be vulnerable, to flag this up as soon as possible, and to seek directions pursuant to rule 3A at the earliest opportunity. Given the difficulties with listing, with which we are all familiar, if a separate hearing is required, it should be applied for as soon as possible. If no separate hearing is required, it will nonetheless be necessary to ensure that there is sufficient time on the day for any relevant issues to be determined. In my view, this responsibility has very real, day-to-day implications. It’s quite likely, if we think about it, that each of us is dealing at any given time with at least one case in which at least one of the parties may be struggling to communicate, and we should consider in each such case whether the court should be invited to assess whether that party is vulnerable. (Ideally, of course, the case would not be in litigation in the first place, but I wonder how many cases proceed into litigation as a result of the party’s inability to communicate as effectively as others might.)

HHJ Greensmith’s analysis

The judge in W v H was really between a rock and a hard place: on the one had he was faced with two parties who hadn’t lived together for a number of years because one of them couldn’t tolerate the behaviour of the other – which, to any normal person (ie not a lawyer) looks like a marriage which has irretrievably broken down. On the other hand, he was faced with this wife who couldn’t support her own evidence and who had always known what she was getting herself into when she married this husband.

The wife, it seems, was incapable in the witness box of backing up her own written evidence. For example, when asked about the husband’s alleged bullying, she said that she did not feel bullied by him “in general” (para 73), and when asked about the husband allegedly forcing her to give him control of all her income, she accepted that she had been able to buy a caravan from her own funds.

By contrast, the judge found the husband to be a credible witness. With one notable exception (to which I refer below), he did not appear to accept that there had been any behaviour on the part of the husband, let alone any behaviour which the wife could not reasonably be expected to live with. Any behaviour there may have been, he seems to suggest, was actually hearsay, exaggerated, confused or even just an “impression” gained by the wife, which she “should have” interpreted in the context of the husband’s autism.

Simultaneously, however, the judge found that the husband’s autism was likely to have had a significant impact upon his behaviour throughout the marriage (and before).

The judge felt that the timing of the wife’s knowledge of the husband’s diagnosis was relevant. That she knew about the diagnosis before her amended petition was drafted was important, he thought. He found that she had sought to capitalise on the husband’s autistic traits in order to satisfy the statutory test. That was not to say that any type of behaviour from someone who is autistic has to be tolerated, but that this wife could reasonably be expected to show a level of empathy and professional understanding of the personality of this husband, whom she had chosen to marry.

I think that the judge may be conflating two issues here: (1) the relevance of the wife’s knowledge of the husband’s diagnosis; and (2) the extent to which the wife knew about the husband’s behaviour before she married him.

As to the first issue, the judge found that the wife had capitalised on the diagnosis. If “capitalising” in this context means that the wife used the diagnosis as evidence to support her claims, then to the extent that such claims were exaggerated, knowledge or otherwise is irrelevant; it is the exaggeration itself that is wrong (“exaggeration” in this context being facts which are not made out). As such, knowledge of the diagnosis may not be relevant.

As to the second issue, section 2(3) of the Matrimonial Causes Act 1973 says that the court must disregard any behaviour which was known to the petitioner and the petitioner nonetheless went on to live, or carried on living, with the respondent. According to this husband, the wife was well aware of his behaviour prior to the marriage (because he had always been like this) and nonetheless went on to live with him. How could it be right or fair, the husband suggested, for the wife to rely on behaviour which had been there all along, behaviour which she had accepted in happier times, but which she now claimed to be so bad that no one reasonable would expect her to live with him?

The judge took a slightly different view. He said that section 2(3) was of “limited application as many of the allegations are traits of behaviour which are continuous in their nature”; in other words, the wife could not be accused of continuing to live with a behaviour. Rather, there was new behaviour all the time, and the question was whether the new behaviour satisfied the test.

The husband did not therefore apparently succeed in that part of his defence. The fact that he had always behaved in this way was not relevant. What appeared to be relevant was that he behaved in this way because he was autistic, and the wife could reasonably be expected to show more empathy and professional understanding towards him.

I referred above to one example of behaviour on the part of the husband which did cause the judge some concern. That was the husband’s refusal to accept the validity of the marriage. The parties had been married in Australia, and the wife had been unable to get hold of an original marriage certificate. She was forced by the husband to go to great trouble and expense to get hold of one, although all along he knew that the marriage was valid. Even in his written statement of evidence and during his oral evidence, he repeatedly described the marriage as “purported”. The judge could see that this might cause significant distress to the wife, but there was no evidence that in fact it had. The behaviour did not therefore satisfy the subjective part of the test.

In summary, it seems to me that although the husband’s autism was highly relevant to the conduct of the final hearing, it may not have been relevant to the outcome, either in terms of his behaviour, or in relation to the timing of his diagnosis. The wife’s evidence was simply so poor that the judge was unable to find in her favour, autism or no autism.

It does however raise questions for us to consider in relation to similar cases. Not divorce cases – behaviour is no longer relevant, save in relation to the diminishing number of legacy cases – but in relation to other cases where at least one of the parties is neurodivergent. Its relevance may be in terms of that party’s conduct.

Autism as conduct

There is a kind of epilogue to this story, contained in that part of the judgment which deals with costs (paras 88–92). The judge found two instances of conduct on the part of the husband which he took into account when making no order for costs against the unsuccessful wife. Those were:

  • The husband’s persistent refusal to accept the validity of the marriage, despite knowing all along that the marriage was valid.
  • An open offer made by the husband was not a genuine attempt to resolve matters. On the contrary, it was an attempt by him to further engage the parties in continuing protracted and expensive litigation. Further, it was an attempt by the husband to place inappropriate pressure on the wife.

I wonder how the wife felt on hearing these findings. Some people might wonder whether she might have succeeded if the court had heard evidence as to the effect on her of these two specific examples of the husband’s conduct.

I might go further:

  • The husband was pursuing his defence, long after no-fault divorce became a thing in April 2022, five years after the wife had left, and three years after she had pleaded that it was no longer reasonable for her to continue to live with him.
  • He claimed in his answer to have “considered” the matter and concluded that it was better for them to remain married notwithstanding the wife’s feelings.
  • He had submitted no fewer than 180 written questions to the wife, of which the judge himself disallowed 76 (for comparison, the wife submitted 30 questions, of which only three were disallowed) – an onerous approach.

All of this might have been evidence of “behaviour” which, if the wife was unable to tolerate living with it and had given evidence to that effect, might have persuaded the judge that there was behaviour on the part of the husband which she should not reasonably be expected to live with.

But regardless of whether the wife could or should have succeeded, it appears that a party who behaves badly in litigation cannot escape liability in costs by pleading autism. That is definitely something we all need to consider and take active steps to manage where relevant.


At the end of the four days in court, the parties were still married. Thankfully, from the wife’s point of view, she was able immediately to apply for a divorce under the new no-fault law, and may well be divorced by now.

Drawing some of the strands together, it seems to me that the importance of this case lies not in its consideration of the old divorce law which, as noted above, is lumbering towards obsolescence, but in its consideration of:

  • The husband’s vulnerability and the importance for all of us to recognise and deal appropriately with a client who may be vulnerable in this way.
  • The ways in which the client may be supported in seeking justice through their status as a vulnerable witness.
  • The vulnerability of a vulnerable witness to liability in costs for behaviour which at least in part they may not be able to control.