Dr Matthew Garrett v Dr Roy Schestowitz & Anor - Find Case Law

This is a dispute between prominent ‘free software movement’ activists. The free software movement advances a philosophy and practice which values the freedom of users to create and share software enabling internet access, and challenges the dominance of ‘big tech’ software and systems over the online experience. That includes a preference for internet relay chat (‘IRC’), an online instant messaging system dating in origin from the 1990s, over the big social media platforms. The challenge the free software movement makes is not only of a technical, but also of a social, economic or ethical nature, and it espouses some wider sets of values accordingly.

The Claimant, Dr Matthew Garrett, describes himself as a software engineer, computer security expert, and free software activist, with an academic background in (computational) genetics. He gained his doctorate at Cambridge University, and has worked with ARM Ltd, a Cambridge-based software design company, throughout his career. He is currently based in ‘Silicon Valley’ in California, where he holds a senior position at NVIDIA, one of the top half dozen big tech companies globally, and a part-time lectureship at the University of California, Berkeley, teaching operating systems security for a masters course. He sets out that he has worked on some of the most significant projects in the free software movement and won a Free Software Foundation award in 2013 for his work on Secure Boot, UEFI and the Linux kernel (at the time, he states, the FSF was one of the most important organisations in the world for the advancement of free software).

The Defendants are a Manchester-based married couple, each of whom operates a UK website focusing on free software information, interests and campaigns, and hosting an IRC channel. Dr Roy Schestowitz describes himself as a software engineer, interdisciplinary researcher and fair competition advocate, with an academic background in medical biophysics. He operates the website www.techrights.org (‘Techrights’). Mrs Rianne Schestowitz describes herself as a computer scientist who works for a FOSS (free and open-source software) specialist. She operates the website news.tuxmachines.org (‘Tuxmachines’). Both websites have been in operation for some two decades. These websites are declared by Dr Schestowitz to be not-for-profit undertakings – they are, he says, a committed labour of love on which they both publish tirelessly to the extent of thousands of articles a year.

Dr Garrett brings a claim in libel over a series of 24 publications appearing on either Techrights or Tuxmachines or both, in August and September 2023. The publications complained of appear at Annex A to this judgment. Dr Garrett brings a further claim in data protection in respect of these and a number of other publications appearing on the websites at around the same time. Dr and Mrs Schestowitz counterclaim in harassment.

Dr Garrett issued proceedings, and filed and served his particulars of claim, in April 2024. This had been preceded, as is usual, by a certain amount of correspondence before action, in which Dr Garrett had requested the removal of the publications from the websites. Dr and Mrs Schestowitz declined to do so.

Dr and Mrs Schestowitz filed and served a defence and counterclaim in September 2024. The libel claim was resisted on the grounds that (a) none of the publications complained of had caused or was likely to cause serious harm to Dr Garrett’s reputation, and (b) one or more of the statutory defences of truth, honest opinion and publication on a matter of public interest applied. The data protection claim was resisted on the ground that the journalistic purposes exemption applied.

Dr and Mrs Schestowitz’s pleadings had been professionally prepared, but they have been acting in person in recent months, including representing themselves at trial. There was a significant amount of interlocutory case management activity, overseen by the Masters, in August and September of 2025, notably in relation to timetabling the provision of witness statements by or on behalf of Dr and Mrs Schestowitz. These had still not been forthcoming by the time of the pre-trial review before Griffiths J on 6^th October 2025.

There appears to be no record that any formal application was subsequently made, although I was shown an unsealed version of an application by Dr and Mrs Schestowitz to set aside the order of Griffiths J. Dr Schestowitz confirmed to me at the opening of the trial on 30^th October 2025, at which he and Mrs Schestowitz were informally supported by a defamation solicitor, that in any event no further order was sought on that (unsealed) application and no further application was being made. He told me that he and Mrs Schestowitz had taken a deliberate decision not to submit any evidence or call any witnesses, on financial grounds. He confirmed that he understood that meant that, while he and Mrs Schestowitz would be able to put Dr Garrett to proof of his claims, to cross-examine the claimant witnesses (Dr Garrett and his solicitor), and to make submissions to me, they would not be able by these means to introduce evidence in support of their defence and counterclaim.

There had been no ‘preliminary issues’ determination ahead of trial in this case. Mr Hamer, Counsel for Dr Garrett, explained that the parties’ pleadings had not been very far apart in this respect. Nevertheless, there was no formal agreement about them.

Legal Framework

Single natural and ordinary meaning

My preliminary task was therefore to “determine the single natural and ordinary meaning of the words complained of, which is the meaning that the hypothetical reasonable reader would understand the words to bear”.

The legal principles I am required to apply in doing so are well established. I directed myself to the useful guidance on determination of ‘single natural and ordinary meaning’ distilled from the authorities (including that of the Supreme Court in Stocker v Stocker [2020] AC 593) and set out in Koutsogiannis v Random House Group[2020] 4 WLR 25,at [11] and [12]. The guidance of the authorities is of course just that – guidance – intended to simplify not complicate the exercise. And each case turns on its own facts. But the following briefly summarises the guidance as it applies to the present case.

The governing principle in determining ‘meaning’ is reasonableness. The intention of the publisher (here, Dr and Mrs Schestowitz respectively) is irrelevant in law: the test focuses on how words are read, not how or why they came to be written. It is an objective, not subjective, test.

Each publication must be considered as a whole, in the context in which an ordinary reasonable reader would have read it. That reference to ‘context’ was explained by Nicklin J in Riley v Murray [2020] EMLR 20 at [15]-[17]. I can, and where relevant should, take account of: (a) matters of common knowledge – facts so well known that, for practical purposes, everybody knows them; (b) matters to be treated as intrinsic to the publication – for example via a hyperlink; and (c) any other material that could reasonably be expected to have been known or read by all the readers of the publication complained of. But otherwise, no evidence beyond the publication complained of is admissible as to what it means. And natural and ordinary meaning does not rely on a reader having any special knowledge.

I am guided away from over-elaborate or lawyerly analysis of text. I need to avoid both literalism, and any strained or forced interpretation. I can and must determine the single meaning I myself consider correct, and am not bound by the meanings advanced by the parties (so long as I do not alight on something more injurious than a claimant's pleaded meaning).

So I am to keep in mind, as guided, the perspective of an ordinary, reasonable reader of each article complained of, reading it once through as it appears, and forming an impression of what it conveys on its face. The reasonable reader is neither naïve nor suspicious; is able to read between the lines and pick up an implication; and is allowed a certain amount of loose thinking without being ‘avid for scandal’.

Fact/opinion

I have further directed myself to Koutsogiannis at [16] and [17], as well as to Millett v Corbyn [2021] EWCA Civ 567, for guidance on considering whether the words complained of contain allegations of fact or expressions of opinion. I am reminded by the authorities that the test for the difference between fact and opinion is an objective one. That comes back to how the words would strike the ordinary reasonable reader. I have to look at the substance, not the intention of the writer or any label the writer may, or may not, have attached.

Subject matter and immediate context can be especially important here. In the classic formulation, “opinion is something which is or can reasonably be inferred to be a deduction, inference, conclusion, criticism, judgment, remark or observation” (Clarke v Norton [1910] VLR 494 at page 499). But sometimes care is needed: there is a difference between comment which is pure opinion and comment which is an imputation of underlying fact.

Factual allegations can be published with varying degrees of certainty as to what is being imputed. These degrees of certainty have been classically analysed at three distinct levels : level 1: the claimant is 'guilty' of the conduct alleged; level 2: there is 'reason to suspect' the claimant is guilty of the conduct alleged; and level 3: there are 'grounds for investigating whether' the claimant is guilty of the conduct alleged (Chase v. News Group Newspapers[2002] EWCA Civ 172).

While ‘meaning’ and ‘fact/opinion’ are distinct preliminary issues, the authorities counsel against trying to resolve them in too linear or compartmentalised a fashion. I have to bear in mind whether this is a case in which the questions of ‘meaning’ and ‘fact/opinion’ might throw light on each other, such that it would be wrong to tackle them in a sequence which proves to be a trap of false logic. I note the risk and seek to avoid it.

Defamatory Tendency

The test at common law for whether a (natural and ordinary) meaning is defamatory is well-established: whether it substantially affects in an adverse manner the attitude of other people towards a claimant, or has a tendency to do so. Some recent authorities put it in terms of identifying that a claimant has breached the common, shared values of our society (Millett v Corbyn). This is not about actual impact at this stage, it is about the meaning of the words themselves and their inherent tendency to damage someone’s reputation. ‘Substantially’ imports a threshold of gravity or seriousness (Thornton v Telegraph Media Group Ltd [2010] EWHC 1414).

Approach, Consideration and Determinations

I adopted the standard approach to the determination of meaning and the other preliminary issues. I read each article complained of quickly once through, before knowing what either party wanted to say about it. I formed and noted my initial impressions. I then read the parties’ pleadings and skeleton arguments or position statements, heard oral submissions (neither Dr nor Mrs Schestowitz chose to address me orally on the preliminary issues), and reserved judgment as to how far to adjust my initial views.

The table attached to this judgment at Annex B sets out the parties’ rival meanings and, in the final column, the matters I have now determined. Where relevant, I have included some short explanatory notes for my decisions where there appeared to be any significant dispute about a particular issue. The following general points also arise.

First, I have had to reflect in a number of respects on whether my preliminary views on meaning needed to be moderated on the ground that they went further than those contended for by the Claimant. Where I have concluded that my meaning was simply a particularisation of matters expressed in more general terms by the Claimant, I have not done so. Where, however, I had identified a meaning not referred to in the Claimant’s pleading, I have withdrawn it for that reason.

Second, while each publication’s meaning is determined on its own terms, some common repetitive themes emerge. Dr Garrett is chiefly accused of an online campaign of material which is (variously) criminal, illegal or offensive. The criminal matters alleged include cybercrime, hate crime, blackmail, issuing threats of violence or death, and matters adjacent to terrorism. Other illegal matters alleged include defamation, harassment and online abuse. Offensive matters alleged include material that is variously racist, antisemitic, misogynist, homophobic or otherwise hateful or discriminatory, sexually incontinent, or drugs-related. Dr Garrett is alleged to have waged this campaign through the medium of IRC ‘sockpuppet’ accounts – accounts under pseudonymous user nicknames intended to be a vehicle for distributing material anonymously and deniably. Many posts from these accounts are reproduced in the articles complained of. Dr Garrett is also repeatedly alleged to be an uncontrolled user of illegal class A drugs, principally crack cocaine.

Third, I have concurred with the parties in understanding that most of the allegations relate to online behaviour. But there are some notable exceptions to that, chief of which is the allegation relating to Dr Garrett’s being an habitual cocaine user.

Fourth, where criminal or illegal acts are alleged, the meanings I have determined are closer to those contended for by the Claimant – namely that these are allegations of fact at Chase level 1. An ordinary reasonable reader would, in my view, have no difficulty in understanding that Dr Garrett is straightforwardly being said to have committed them. Where, however, offensive attitudinal matters are alleged – racism, misogyny, etc – the meanings I have determined are closer to those contended for by the Defendants – namely that these are evaluative expressions of opinion.

Although it is not formally admitted, there can be no doubt at all that each of the publications is of grossly ‘defamatory tendency’ at common law. They allege multiple and serious transgressions of society’s norms, up to and including the criminal law.

Legal Framework

By section 1(1) of the Defamation Act 2013, ‘A statement is not defamatory unless its publication has caused or is likely to cause serious harm to the reputation of the claimant’.

The leading authority on this provision is the decision of the Supreme Court in Lachaux v Independent Print Ltd [2020] AC 612. As Lord Sumption’s judgment makes clear ([12]-[14]), section 1(1) imposed a new threshold test for defamation claims, the application of which is to be determined by reference to the actual facts about the impact of a publication, and not just to the meaning of the words. It requires real-world effects to be established beyond the inherent ‘defamatory tendency’ of any publication. The statutory term ‘has caused’ points to some historic reputational harm, which can be shown actually to have occurred; and ‘is likely to cause’ points to probable, actual, future reputational harm.

The serious harm test is a question of fact, and facts must be established by evidence. Facts and evidence – and causation – are matters which are entirely case-specific. But a claimant’s statutory task of establishing that publishees changed their minds for the worse about them – and to a degree warranting the description of ‘serious reputational harm’ – because of reading something, poses clear evidential challenges. That is partly because of the nature of the harm in question: the ‘harm’ of defamation is the effect of a publication in the mind of a third-party publishee (reader), and not any action they may take as a result (nor is it the direct effect of a publication on a claimant reading it themselves). And it is partly because of simple practical considerations: particularly in cases of mass publication, the minds of the publishees may be effectively evidentially unreachable. In such cases, finding enough readers willing to testify that they thought the worse of a claimant has long been recognised as a prohibitively impractical undertaking.

The authorities have therefore developed some detailed guidelines for trial courts applying the serious harm test in ‘mass publication’ cases. In such cases, Lachaux confirmed (at [21]) that a claimant may be able to discharge their evidential burden of establishing serious reputational harm by inference drawn from a factual matrix combining the meaning of the words, the situation of the claimant, the circumstances of the publication(s) complained of and the inherent circumstantial probabilities.

That matrix, in turn, has to be established as a matter of fact and evidence. But again, the authorities give some assistance. The well-established ‘grapevine’ or ‘percolation’ tendencies (Slipper v BBC[1991] 1 QB 283; Cairns v Modi[2013] 1 WLR 1015) of defamatory publications, particularly online and through social media of any kind, may in an appropriate case be factored into inference about scale of publication. And the likely identity, as well as the numbers, of at least some of a class of publishees may be relevant to the assessment of harm, for example where some individuals may be particularly positioned to lose confidence in a claimant or take adverse action as a result. But these are highly fact-specific matters; the inferences which may properly be drawn in any individual case depend entirely on the circumstances of that case. The Court of Appeal in Blake v Fox[2025] EWCA Civ 1321 has, however, recently indicated that if a claimant establishes a factual matrix of mass publication, grave defamatory tendency of the content, and the claimant’s significant reputational exposure, then the inference of serious reputational harm may have to be considered by a court to be irresistible.

Consideration

Approach, and Gravity of Imputation

Dr Garrett advances a wholly inferential case of serious harm, relying on this being a case of mass publication, very grave inherent defamatory content, the engagement of his own professional and reputational capital, and inherent probabilities. The burden lies on him to establish this underpinning factual matrix.

As a starting point, I have no difficulty in the first place in recognising the heavy gravity of the allegations made in each and every publication complained of. Dr Garrett’s responsibility for a relentless online campaign of harassment and abuse, by the covert means of sockpuppet accounts, is advanced as fact. Extensive, multiple and serious online criminality is advanced as fact. Dr Garrett is accused of publishing a range of toxic or highly offensive online material by reference to direct quotations from the sockpuppet accounts he is accused of authoring; this material is on its face grossly offensive in many dimensions and carries clear implications of the social and moral turpitude of its author. Habitual illegal class A drug use is advanced as fact. In the meanings I have determined – and as to which the parties’ pleaded cases were not in the end separated by a wide margin, since they largely spoke for themselves – the publications complained of are undoubtedly very grave indeed. I did not hear that seriously disputed.

Dr and Mrs Schestowitz however make a double challenge to Dr Garrett’s task of establishing a factual matrix to support an inference of serious harm. They put him to proof (a) that this is in fact a mass publication case, and (b) that his reputation is materially engaged – in each case, within the UK.

The Situation of the Claimant

Taking the latter first, the ‘situation of the claimant’ in this case is that Dr Garrett was born in Ireland and moved to England with his family as a very young child. He emigrated to California in 2009 and has not lived in the UK for 16 years. He holds UK, Irish and US citizenship. To succeed on his claim as brought, he has to demonstrate serious reputational harm within the court’s jurisdiction – that is, in England and Wales. His evidence about his reputation here is as follows.

Dr Garrett states that his reputation in England is both established, and ‘immensely important’ to him. Before 2009 he had worked for the University of Cambridge and had done other Cambridge-based work, starting by doing contract work and working for a tech start-up, before going on to work for Collabora (described as a private open-source software consulting company headquartered in Cambridge) and for Red Hat, a US company with a wholly-owned British subsidiary which had been his employer here.

He draws attention to his current senior role with NVIDIA, which ‘does significant business in Britain’. In that role he has undertaken collaborative work with British companies including Collabora, Linaro, ARM and other consulting firms, including while present in the jurisdiction. He describes these ‘ongoing collaborations’ as being ‘extremely important in terms of providing value to my employer since then’. His evidence is that ‘the reason I am valuable to my employer and my previous employers, is that I am exceedingly good at working across a range of different technologies and with a range of different companies. If my ability to work with any of the companies who are working in a specific field is impaired, the value proposition of me for my employer goes down’. He states that damage to his reputation in England ‘would have an impact on NVIDIA’s willingness to continue employing me’.

Dr Garrett draws attention to the fact that he holds a teaching position at a US university with an international reputation in his field, including in this country, and was awarded a ‘distinguished faculty award’ by that university. His evidence is that his past and present academic reputation is engaged in England. Of his future plans, he states that ‘my goal in the end, at the point where I am in a position to retire, is to be able to move back to the United Kingdom and potentially adopt a position at a university here’.

Dr Garrett’s evidence is that he is regularly invited to speak at major events (with thousands of attendees) concerned with free software, Linux or cybersecurity generally, and has done so in the UK on half a dozen occasions in recent years. He has also spoken at major European events with significant attendance from UK participants, including in the last couple of years.

Dr Garrett also gives evidence that he has many friends and colleagues in the UK.

I did not hear this evidence materially challenged in substance. Dr Garrett’s evidence on the nature, extent and value of his reputation in England and Wales was clear and straightforward. I can and do accept it. It enables me to find as a fact that his professional, employment, academic and personal reputation is materially engaged within the jurisdiction. I make that finding accordingly.

Extent of Publication

The principal dispute in this case on the question of serious reputational harm is, however, that of extent of publication (within the jurisdiction) of the articles complained of. Mr Hamer contended for an inference of a readership in the tens of thousands, within the jurisdiction, for each article complained of, to date. The evidence of this before me was limited. I do not of course have any evidence from Dr and Mrs Schestowitz, and the pre-trial disclosure exercise revealed no analytics for readership of the articles in question (the defence states that ‘site analysis data is wiped after short periods of retention’).

It is not in dispute that the websites get something like a million hits a day. A hit is not an indication of readership; a single page may require a number of ‘hits’ to be viewed. Dr Schestowitz made a challenge that something like half of those could be immediately discounted as machine-generated (bots, crawlers and so on). He made a challenge to the remaining figure based on the proportion of the anglophone world attributable to the UK being a percentage in the low single figures. He made a further challenge based on the high rate with which new articles are posted on both websites, and the relative age (and therefore incremental occlusion) of the material complained of. He challenged that there was no evidential basis in the ‘million a day’ figure for a readership above tens in the UK on any one day, and that many of them would likely be repeat viewers, reading the same articles again and again. Dr Schestowitz pointed out that neither he nor Mrs Schestowitz publishes on, or is a user of, mass social media platforms, that their websites attract viewers with similar preferences against modern social media, and that the grapevine or percolation effects often deployed in defamation litigation to encourage inference of mass publication are accordingly of limited application here.

In these circumstances, and having been put to proof of extent of publication, Dr Garrett’s case that this is an instance of mass publication was advanced evidentially by him, and in argument by Mr Hamer, along the following lines.

First, all of the articles complained of remain online, and have done so since their original publication more than two years ago.

Second, a number of the articles appear reduplicated on both websites, and a number cross refer to each other via embedded hyperlinks. The final publication (the Wiki page) is a compendious hyperlink library of nearly all the articles complained of. The articles, in other words, contain within themselves mechanisms of accessibility and republication, and hence maximise readership potential.

Third, and bearing in mind the extent of his reputational engagement in the UK, recent Google searches on Dr Garrett’s name from a UK location were bringing up links to Techrights and Tuxmachines articles complained of (including the final Wiki page publication) on the first page, and, it appears, also did so around the time of the original publication (as one of the Tuxmachines posts seems to have acknowledged). That suggests both (a) the continuing ready accessibility of the material complained of to those interested in Dr Garrett and (b) the historical popularity of actual access to the Schestowitz websites via Google. Google is of course only one among a number of popular search engines.

Fourth, both websites are well-established. It is conspicuous that each has been active over two decades; that is a significant marker in the field. Dr Garrett’s evidence is that they are serious, sought-after, well-regarded, popular and trusted in the anglophone free software community, including for news, campaigning and educational content. Neither Dr nor Mrs Schestowitz suggested otherwise.

The final piece in the jigsaw Mr Hamer sought to assemble is a centuries-old English legal principle, derived from the case of Armory v Delamirie(1721) 93 ER 664, that, to the extent that there are gaps in trial evidence, a party who is responsible for them is not entitled to benefit from them. The principle in Armory has been recognised in the caselaw as having a potential to apply in defamation actions to assist a claimant in discharging their burden of proof, where a defendant has failed to provide disclosure or evidence of extent of publication (see, for example, Dudley v Phillips [2022] EWHC 930 (QB) at [25]). Here, Mr Hamer submitted, not only were the relevant analytical data not preserved or disclosed, but a deliberate decision was made by the Defendants to provide no evidence of the extent of the publications for which they were responsible, and therefore not to submit themselves to cross-examination about it. In those circumstances, Mr Harmer submitted, readership of at least tens of thousands could fairly be inferred in relation to any and all of the publications complained of.

I have reflected carefully on Dr Schestowitz’s challenge, Mr Hamer’s submissions and Dr Garrett’s evidence. I bear in mind the years-long persistence, cross-referencing and indexing apparent within the articles. I can and do place weight on the evidence I have to suggest that the publications complained of were accessible to, and accessed by, a UK public searching online for information about Dr Garrett. It is common ground that Techrights and Tuxmachines are popular and authoritative sites, including in the UK, among those interested or active in the free software community. I do not understand it to be in dispute that that wider community in the UK is otherwise than substantial, perhaps in the tens or hundreds of thousands. Dr Garrett is, I have accepted, someone with a profile in that community, and the allegations in the articles complained of are of a sensational and eye-catching nature (and tone). In these circumstances, it appears to me that a total readership of the articles to date, including by way of some onward republication (percolation), of the order of at least thousands is more probably than not a fair estimate. And if there are doubts about that, Dr Garrett is entitled to rely on the Armory principle to have them resolved in his favour.

Inherent Probabilities and Conclusions

A readership of at least thousands makes a fair case, in context, for regarding this as an example of ‘mass publication’. It is not of the order of a national newspaper or news channel. But nor is it a case dealing with publication to a limited class of identified or identifiable individuals. Just as important as the raw numbers is the identity of the likely readership. I am satisfied that publication in this case was to precisely the constituency – individuals in the UK interested in the free software community and in the doings of free software activists, and who had either already heard of or were likely to take an interest in Dr Garrett – with which Dr Garrett’s UK reputation was most closely engaged. I have to consider that in conjunction with the gravity of the allegations. Sockpuppetry, trolling, inappropriate sexual content and hate speech are allegations of profound breaches of the ‘netiquette’ of this community in their own right. The accusations about the content of this behaviour – serious online crime – push those allegations to the extremes of gravity in context. Coupled with the imputations going to Dr Garrett’s offline private life, including sexual misconduct (or crime) and the habitual use of illegal class A drugs, it is hard to imagine a more serious assault on the professional and personal reputation of a prominent free software activist.

In all these circumstances, I accept Dr Garrett’s case that not only is the inherent meaning of the publications complained of grave, but they have caused and are likely to cause him serious reputational harm. Indeed, on the basis of the factual matrix advanced and established, and mindful of the guidance in Blake v Fox, the inference of serious harm may have to be regarded as inescapable.

Legal Framework

All three defences are pleaded in the present case.

The Truth Defence

Pleading and evidence

The pleading of the truth defence – that is, the pleading that what is said about Dr Garrett in the publications complained of, in the natural and ordinary meanings I have determined, is in objective fact true – relies on (a) a small number of messages Dr Garrett accepts he posted and (b) his asserted authorship of a campaign of threats, abuse and harassment conducted from a list of named sockpuppet accounts.

The authorship or control of these accounts has consistently been strenuously denied by Dr Garrett. I have no evidence from the Defendants to support it. Instead, they necessarily rely on an inferential case built on a limited number of pleaded facts, some of which are undisputed. I consider them in turn.

First, the defence relies on an incident a few years prior to the sockpuppet campaign, in which Dr Garrett admittedly registered himself online in two borrowed names for the purpose of making a rhetorical or satirical point about the owner of those names. Dr Garrett’s evidence is that he did so openly, and with the knowledge of the owner (who had himself vacated the names in order to borrow another user’s nickname – which was in part Dr Garrett’s point in also doing so). Dr Schestowitz clearly takes exception to that particular piece of theatricality as a major breach of netiquette in its own right, and regards it as a sign or symptom of propensity for sockpuppetry. But the incident in question, on its face, is plainly something quite different from the covert use of sockpuppet accounts to publish illegal or offensive material, and is not in my judgment capable of indicating any sort of propensity to do that.

Second, it is said there was an incident in which Dr Garrett’s own named account and one of the sockpuppet accounts experienced simultaneous dropped connections. If established, that could indicate dual operation by a single individual. The evidence from the Claimant is that the dropped connections were not in fact simultaneous. I have no evidential basis for doing otherwise than proceeding on that basis. This pleading cannot in the circumstances support an inference of Dr Garrett’s authorship of the sockpuppet posts.

Third, there are some facts pleaded which are capable of linking the sockpuppets to each other. They are not capable of being linked to Dr Garrett by that means.

Fourth, it is pleaded that after Dr Schestowitz had made a complaint naming Dr Garrett as the suspected author of the complaints, they subsided. I have no evidence of this – of its extent, timing, or at all.

Fifth, it is suggested there is some similarity in the language used by Dr Garrett in his own account and that in one of the sockpuppet accounts. That suggested similarity is not particularised. I was not taken to it by Dr or Mrs Schestowitz. It is not immediately obvious on the face of the publications complained of themselves.

Sixth, one of the sockpuppet accounts on one occasion posted a short insult in the Irish language. Dr Garrett’s evidence is that he does not speak Irish and would have been incapable of posting that insult.

Seventh, it is suggested that one of the sockpuppet accounts used the first person singular in referring to an accusation made against Dr Garrett. I was not taken to this example, or its context.

Dr and Mrs Schestowitz took me to a small number of posts Dr Garrett had made on a website he had operated some 25 years ago, when he would have been aged about 20. In one, he made a comment about people feeling the need to punch someone. In another he referred to being sorely tempted to write a book entitled Axe-murdering for fun and profit as a disincentive to people bothering him. In a third he expressed exasperation with some individuals along lines that it made him want to stab them and then stab himself. In a later social media post in 2018 on Twitter (X), he was involved in a game of listing criteria for electing a new prime minister, in which he proposed ‘1. Cancel Brexit; 2. Decriminalise most drugs; 3. Spend less on candles [an internet meme]; 4. Resign’.

I had no difficulty in immediately recognising all of these as examples of jocular or facetious rhetoric. Dr Garrett’s own evidence was that he now considered them callow and in poor taste; they were mostly decades old, and he would not express himself in this kind of way any longer. Dr Schestowitz plainly regarded them as unpleasant breaches of netiquette. But the proposition that these examples should be taken literally, and this sort of online discourse considered indicative of a propensity in real life to the use of sockpuppetry to threaten or commit violence, or to an illegal drugs habit, cannot begin to get off the ground. They are clearly nowhere near any such thing. Nor are such allusions to bladed weaponry as they contain any sort of basis for inferring that sockpuppet accounts under names such as ‘Dr Axe’ were being operated by Dr Garrett.

It is Dr Garrett’s evidence that at least some of the sockpuppet posts were posted at a time when he can establish he was teaching a university class. It is also his evidence that it was not practically possible for him to have been the author of those posts, both because he did not have access to the IRC channel on his phone and because he could not have done so without attracting the attention of his students.

Assessment and conclusions

It is plain that the onslaught of sockpuppet trolling to which Mrs Schestowitz in particular was subjected was a truly appalling experience – the very antithesis of everything Tuxmachines and Techrights stands for in the free software community. There is not the slightest suggestion that either Dr or Mrs Schestowitz did anything to invite or deserve it. They are both clearly and justifiably angered, dismayed, distressed and hurt by it.

It is also plain that Dr and Mrs Schestowitz do not think highly of Dr Garrett in general. Some of that appears to have had its origins in differences of professional views. Some of it may be to do with his choices to move to Silicon Valley and to work for, with or alongside big tech companies; Dr Schestowitz explained clearly to me how and why they themselves have made different choices. Some of it may be to do with Dr Garrett’s modes of online expression and what they consider to be his poor netiquette; Dr and Mrs Schestowitz explained to me that these are matters they take very seriously, and that they are serious people running serious websites dealing in serious technical and socio-political matters. They do not admire facetiousness.

It is also plain enough that Dr and Mrs Schestowitz have found it entirely straightforward to convince themselves that Dr Garrett was behind the trolling. But the task they have given themselves in pleading the truth defence in defamation proceedings is to establish or prove, on the balance of probabilities, that that is objectively true. And they have made that exceptionally difficult for themselves by advancing no evidence for it. I can only uphold a truth defence if I am given a sufficient evidential basis for doing so. I have looked hard at the defence as pleaded. I can recognise the matters I have been shown, in all the contextual circumstances, as the reasons Dr and Mrs Schestowitz advance to explain their suspicions of Dr Garrett. But I cannot possibly recognise in them, singly or together, anything capable of establishing the objective truth that Dr Garrett was their troll. They are simply not capable of sustaining the weight of a truth defence, or enabling me properly to find that, more probably than not, what the publications said about Dr Garrett was the verifiable truth.

To make such a finding, I would have both to reject Dr Garrett’s own clear evidence, and to discount the possibility that, if Dr and Mrs Schestowitz suspected Dr Garrett of the posts, some third party operator(s) responsible for the sockpuppets might at some level have encouraged that.

I found Dr Garrett in general a straightforward witness who gave a clear account of himself under sustained challenge in the witness box. I have been given no proper basis for rejecting his plain, unqualified and vehement denial that he ever had anything to do with the sockpuppet campaign against Dr and Mrs Schestowitz. The matters pleaded by way of the truth defence barely raise a statable case that he did. I accept Dr Garrett’s evidence. He is entitled in these circumstances to a clear finding that I have not been shown any evidence on which I can fairly place any weight, that he was responsible for the sockpuppets.

That being so, there is no such evidence before me that Dr Garrett habitually uses illegal class A drugs. There is no such evidence he has committed any crimes. There is no such evidence he has threatened, abused or harassed anyone, or that he has espoused or promulgated repugnant and hateful views. There is no basis on which I can properly find any of the many factual accusations levelled against him in the publications complained of to be true. The truth defence must fail in its entirety.

The Honest Opinion Defence

The honest opinion defence is pleaded bare, otherwise than by reference to the allegation that Dr Garrett was responsible for the sockpuppet trolling. In the meanings I have determined, the allegations identified as matters of opinion are all identified as descriptors of the content of the campaign he is alleged to have pursued. That is the only ‘basis of opinion’ articulated in the articles complained of for the purposes of section 3(3) of the Defamation Act 2013.

If, as I have concluded, I am unable to uphold the contended factual truth of Dr Garrett’s responsibility for the sockpuppet trolling, that is not a basis on which the honest opinion defence can be upheld either. An honest person has no basis in contemporaneous fact, as required by section 3(4)(a), for an opinion that Dr Garrett was responsible for an offensive and hateful campaign, when he cannot be shown to have been responsible for the campaign at all. The honest opinion defence is unsustainable accordingly.

The Public Interest Defence

I have reminded myself of the review of the authorities on this defence set out by Nicklin J in Turley v Unite The Union [2019] EWHC 3547 (QB) at [137]-[138]. A three stage approach has to be taken. Stage one asks whether the statement complained of was, or formed part of, a statement on a matter of public interest. The second establishes whether the defendant believed publishing the statement complained of was in the public interest. The third asks whether any such belief was reasonable.

The pleading of the defence in the present case addresses the first question by asserting that all the statements complained of were publications on a matter of public interest, namely ‘the exposure of a vicious and persistent troll committing, among other wrongs, communication and harassment offences in an attempt by the Claimant to intimidate and obstruct promotion of computing standards argued for by the Defendants and that have wide significance within the field of computing’.

The authorities guide that public interest is a broad concept in this context, and can be considered in the light of whether publications make a contribution to a matter of public debate, but is distinguishable from matters that are personal or private. I can see that the promotion of high standards of online behaviour, and the calling out of unlawful, intimidatory or suppressive behaviour, are capable of being included in the broad ambit of a ‘matter of public interest’. The question here would really be about whether any of the publications complained of is recognisable as contributing to it. Dr and Mrs Schestowitz took a decision not to write general articles about the trolling they had received (nor, indeed, it would appear, to block the sockpuppet accounts), but instead (a) to republish quantities of it and (b) to attribute it to Dr Garrett in strongly and personally deprecatory terms. There can be little doubt on a fair reading of the articles that their focus is squarely on attacking Dr Garrett rather than any more objective or issues-based purpose.

I have no evidence as to any due diligence undertaken before the articles complained of were published. The defence is pleaded by reference to the reasonableness of reliance on the assemblage of matters I have considered in relation to the truth defence above, which, for the reasons I have given, are not reasonably recognisable as an exercise in considered research and verification. Dr Garrett was not approached to give his side of the story. One of the articles complained of (publication 14 and 15) reproduced a paragraph from a letter Dr Garrett had written on 5^th August 2023, stating that ‘As I have previously informed you via email, I am not responsible for the elusive_woman IRC account or any of the other accounts you have incorrectly asserted I am linked to. I have provided evidence that I am not that person’ – namely that one of the postings from the account mentioned was made at a time he was teaching a class. The letter had gone on to challenge Dr Schestowitz for appearing to accept that others might have been involved in the trolling, but that part of the letter was not reproduced. Beneath the (cropped) picture of the letter was commentary beginning ‘Oh, wow! How will I know if you are lying again?’ and continuing with invective and multiple accusations against him. This has little claim to being in the nature of affording a fair right of reply.

Dr Schestowitz made an eloquent case to me for the websites he and his wife operate being quintessentially public interest, not-for-profit endeavours, conscientiously dedicated to the interests of the free software community, trusted and well-regarded as such. I can accept that. I can also easily understand that he and Mrs Schestowitz had felt attacked and threatened by the sockpuppet trolling, on both a personal basis and on behalf of all they were trying to do and achieve with their websites. That is not, however, enough to engage the public interest defence to publication of libels that cannot be shown to be true, blaming and defaming someone for the trolling on no better basis than that produced in these proceedings. Any belief in the public interest in naming and shaming Dr Garrett – repeatedly, and in the most personal and derogatory terms – on such a basis could not in my judgment be regarded as a reasonable one for the purposes of the defence.

In all these circumstances, the public interest defence inevitably fails.

Dr and Mrs Schestowitz are proud of Techrights and Tuxmachines – their lives’ dedicated labour, passion and achievement. They are proud of the websites’ decades-long record of hosting an important and well-regarded IRC for the free software community in the UK and beyond, and of the thousands of informed and informative articles the websites have carried, including the results of some significant investigative journalism.

They were understandably horrified to become the victims of an anonymous internet campaign of abuse, threats and harassment a couple of years ago. Nothing could have been further from the websites’ ethos, and it must have been traumatising for them on a personal level. Their response was publicly and repeatedly to blame Dr Garrett for it all, having convinced themselves they had some reasons for suspecting him of being behind the anonymous trolling – and to do so in a manner that was bitter, highly personal and thoroughly derogatory. They were undeterred by his protests that he had nothing to do with it. They steadfastly refused to take their accusatory material down.

Dr Garrett has been successful in advancing a defamation claim to the point at which Dr and Mrs Schestowitz are required by law either to concede the claim or defend it. They chose to defend it, but they chose to do so without advancing any evidence in support of their defences. A court’s job is to apply the law to the facts as it finds them, and to find the relevant facts according to the evidence before it. If it is provided with no evidence in support of a libel defence, there is a limit to what a court can do.

I have considered the defences pleaded by Dr and Mrs Schestowitz as best I can, given that they provided me with no evidence to work with. The bases they advanced for their suspicions that Dr Garrett was behind the sockpuppet trolling were simply not capable in the circumstances of adding up to an evidential proof on which a court could properly find that was, objectively, true. That allegation was at the heart of their resistance to Dr Garrett’s claim, and they have not enabled me properly to do anything other than reject it as capable of founding a defence, either by establishing that they had done due diligence at the time, or by advancing evidence or proof of objective truth I could consider at trial.

A defamation claimant in these circumstances is entitled to a judgment, and to remedies, which repudiate the publications complained of, vindicate his reputation, and restore his good name.

Mr Hamer indicated to me at trial that, in the event that I upheld Dr Garrett’s claim in defamation, I could consider myself invited to conclude that it was unnecessary for me to consider the data protection claim on its merits. In the circumstances, I accept that invitation. Although the data protection claim extends to a number of publications additional to those on which the libel claim was based, and although it extends to the processing of Dr Garrett’s personal data in forms other than publication, I am satisfied that embarking on a full analysis of the data protection claim would be disproportionate to any additional vindicatory justice, or remedies, it could potentially achieve for Dr Garrett.

Dr and Mrs Schestowitz’s counterclaim is pleaded by reference to Dr Garrett’s alleged responsibility for the sockpuppet course of conduct. For the reasons I have already given, I cannot make the necessary findings of fact on which that allegation inevitably relies. The counterclaim as pleaded necessarily fails for that reason.

At the trial hearing, Dr and Mrs Schestowitz also made extensive and sustained submissions objecting to the manner in which Dr Garrett and his legal team have been conducting this litigation against them. It is plain that they have both, in general, found the experience of being proceeded against for libel intrusive, frightening and intimidatory. I was told that they experienced it as motivated by an intention to harm them and their websites financially, and to interfere with their investigative journalism. They objected specifically to Dr Garrett’s having made contact with their internet service provider and other web hosts in an attempt to have the publications taken down, to the obtaining of their postal and online contact details, to the threatening tone of solicitors’ correspondence, to the conduct of the disclosure process, to the conduct by the same firm of solicitors of a second set of proceedings against them which Dr and Mrs Schestowitz had unsuccessfully sought to have consolidated with Dr Garrett’s claim, and to a reference made in formal documentation to a previous name of Mrs Schestowitz, a matter to which she took extremely strong exception. They felt, in short, that they had been ‘harassed’ by this litigation.

I emphasise that none of this formed any part of their counterclaim as pleaded, nor is it evidenced, and I am not as a result in any proper position to make any findings of liability in harassment based on it. I also had to explain the constraints of legal professional privilege on some of the matters Dr and Mrs Schestowitz wanted to explore with Dr Garrett and his solicitor under cross-examination, in relation to their counterclaim.

I have to record also that Dr Garrett in turn takes strong exception to the manner in which Dr Schestowitz has himself conducted this litigation as a litigant in person. Mr Hamer referred to what he considered to be racist attacks on Dr Garrett’s lawyers, posted on Techrights, which he described as probably the worst example he had seen of such conduct.

All of this may need to be explored further if and when it comes to the stage of dealing with the costs of this litigation. That is a matter on which the parties will need to reflect following hand-down of this judgment. It may be that formal submissions and witness evidence may in due course be needed to resolve it. But none of it is properly relevant to the determination of liability on the counterclaim as pleaded.

For the reasons I have given, I cannot be satisfied that, more probably than not, Dr Garrett was responsible for the course of conduct comprising the sockpuppet trolling. The counterclaim must fail accordingly.

Legal Framework

Libel damages

The starting point for any award of damages in any tort is that the court should award a sum of money which will, as nearly as possible, restore the injured party to the same position he would have been in had the tort never been committed.

But the nature and history of libel, and its focus on reputational harm and the restorative vindication of reputation, has resulted in the evolution of a distinctive approach to assessing damages for defamation. As Nicklin J observed in Monir v Wood [2018] EWHC 3525 QB at [228], ‘Damages for libel cannot be calculated on any mathematical basis. By definition, they seek to provide compensation for harm that it is almost impossible to quantify in monetary terms’. The exercise is necessarily therefore a broad and holistic one.

As well as being broad and holistic, the exercise is also highly fact-specific. But the framework of principle does include maintaining broad comparability within the tort and between torts. So regard may be had to other awards in defamation cases of a comparable nature (although the authorities emphasise that no two defamation cases are ever really the same). Regard may also be had to the (very differently assessed) awards in personal injury cases to ensure that damages for defamation are, and are seen to be, proportionate and realistic. But these comparative exercises are by way of guidance only, and the focus must remain, intensely, on the circumstances of the individual case.

The relevant legal principles particularly applicable to assessing libel damages were set out clearly by Warby J (as he then was) in Sloutsker v Romanova[2015] EWHC 2053 at [74]-[82], and in Barron v Vines [2016] EWHC 1226 at [20]-[21]. Broadly, the purpose of an award of damages in defamation proceedings is to compensate for injury to reputation and to feelings, and in particular to vindicate claimants, so far as money can to do that. Vindication and compensation are not to be thought of in compartmentalised terms: the overall purpose of the award remains to restore a claimant, to the extent money can do so, to the position as if the libel had not occurred. But in defamation cases that means not only redressing the balance in terms of quantifiable losses, but unequivocally, albeit proportionately, restoring a claimant’s standing to its previous state. As the authorities put it, the sum awarded must be an outward and visible sign of vindication, sending a message restoring a claimant’s good name ‘sufficient to convince a bystander of the baselessness of the charge’. If an award fails to achieve vindication, it fails properly to compensate and restore the status quo ante.

In assessing damages, the court takes account of all the relevant facts, and in particular the gravity of the defamation, the extent of its publication, and evidence of the harm it has done. The award must in the end be no more than is justified by the legitimate aim of protecting reputation, necessary in a democratic society in pursuit of that aim, and proportionate to that need.

Injunctive relief

Injunctive relief is regularly afforded to successful defamation claimants, to restrain the publication of the same or similar libel. Final injunctive relief, restraining publication post-judgment, is distinctively different from pre-trial interim relief, since the implication of the judgment is necessarily that the published libel is not a species of free speech protected by Article 10 of the European Convention on Human Rights. Nicklin J’s observation about permanent, post-judgment, injunction in Lachauxat [239] was that ‘it is the natural remedy that flows from the Court’s decision’, and, in the absence of satisfactory undertakings, it may be necessary to injunct further publication to give full effect to that decision.

But, importantly, it remains a discretionary remedy. It is not available as of right. In considering whether to exercise its discretion, and if so how, a court will have regard to all the circumstances of a case, including the conduct of the parties, and will focus in particular on assessing the risk of repetition of the defamation should injunctive relief not be granted.

Order for publication of judgment summary

As confirmed by Monir v Wood at [239]-[242], the power to order publication of a judgment summary is itself a distinct incursion into a defendant’s rights to freedom of expression, protected by Article 10 of the ECHR, and must be exercised mindfully of that. The incursion involved must be justified by reference to the legitimate aim of protecting reputation, and necessary and proportionate to that aim.

Quantum of damages

The most significant relevant factors in the present case seem to me to be these.

First, there is the gravity of the libellous allegations in this case which, in multiple and detailed respects, profoundly impugn Dr Garrett’s personal integrity, social standing and professional credibility.

Second, there is the fact that they were published from websites regarded as authoritative and reliable, and repeatedly, to the very audience – including the free software community in England – in which Dr Garrett’s reputation here is most deeply invested, and which has the highest capacity to think the worse of him as a result. This was a substantial audience estimated to be of the order of at least thousands.

Third, there is the fact that Dr and Mrs Schestowitz have insisted on the truth of their allegations all the way to trial, have offered no evidence for it, have refused to take down the material complained of and have made no retraction of or apology for the content. They appear to have thought that this was a point of journalistic principle and integrity, but they do not appear to have engaged in that capacity with the need or desirability for doing more, when faced with Dr Garrett’s assurances that he had nothing to do with the sockpuppet accounts, than simply dismissing them as probable lies. Mrs Schestowitz did however, in her closing submissions at trial, offer Dr Garrett an apology to the extent of having hurt his feelings. I thought that was at least well-intentioned and genuine, and I give her credit for it. I note also that, while Dr Schestowitz challenged Dr Garrett hard in a range of matters under cross-examination, he did so courteously and without directly maintaining the underlying truth of the content of the allegations beyond those of his responsibility for the trolling.

Fourth, I note Dr Garrett’s evidence that the libels have caused him to fear for his professional and academic reputation, for the security of his employment and for his future employment prospects. He has had experience of being approached directly by a colleague and a former colleague over the content of the publications and had to defend himself, and has suffered the natural anxiety and humiliation of wondering what others are thinking of him and whether he may be being avoided or overlooked for opportunities because of what has been said about him. I bear in mind also that he does not suggest there is evidence to date that he has suffered actual employment detriment, and I note the strength of his professional reputation and the extent to which that might be able in itself to be robust to the depredations of gossip and rumour.

Mr Hamer asks for a single global sum to vindicate Dr Garrett’s reputation and compensate him for distress in relation to all the publications complained of. He proposed a range of comparator decisions for my consideration, in support of a submission that libel damages approaching £100,000 would be appropriate. I have considered these. I noted in particular the case of Fentiman v Marsh [2019] EWHC 2099 in which an award of £55,000 was made in respect of allegations in a blog read by about 500 people that the claimant, a company CEO, was a hacker responsible for illegal cyber-attacks on a company. The tone of the allegations there were something comparable to those in the present case – somewhat personally and floridly put. I hold the effects of inflation in mind.

In my judgment, in all these circumstances, the minimum sum necessary to convince a fair-minded bystander of the baselessness of the allegations against him, to vindicate his reputation and restore his standing, and to compensate him for the consequences he has suffered, is £70,000.

Injunctive Relief

Dr and Mrs Schestowitz continue to publish and republish the libels. They have no entitlement to do so. This is not a form of freedom of expression which is protected by law. They have published further similar defamatory statements since the commencement of these proceedings. In the absence of any indication that they will voluntarily desist, remove the material objected to, and give satisfactory undertakings not to repeat the same or similar allegations, they face being compelled to do so by means of an injunction enforceable by proceedings for contempt of court.

A principal consideration in deciding whether to exercise the discretion to compel publication of a summary of a libel judgment is whether it is likely to be effective in coming to the attention of those to whom the original libel was published, and therefore in undoing its original reputational harm and vindicating the claimant.

Publication of a short summary of this judgment on Techrights and Tuxmachines would, in my view, be likely to be effective in reaching the UK audience to which the libels were published. It would be a clear statement to the free software community here that Dr Garrett cannot be regarded as having any stain on his character arising out of the accusations carried by those websites and complained of in these proceedings. That would, in my judgment, be a necessary and proportionate means of vindicating and protecting his reputation and rights.

Dr Garrett’s claim in defamation succeeds. He is entitled to the vindication of this judgment, and to the remedies I have indicated. The counterclaim of Dr and Mrs Schestowitz in harassment is dismissed.

Dr and Mrs Schestowitz experienced appalling anonymous abuse, threats and harassment online. They reacted by using their websites – well-established, respected and trusted in the free software community, and carrying their own personal authority as free software activists – to blame and upbraid Dr Garrett for being responsible for this, and to vilify him accordingly. They have been unable to establish any recognisable basis in law for doing so. They have advanced no evidence for it; I was shown nothing in these proceedings capable of establishing in a court of law that Dr Garrett was in fact responsible or had anything to do with it. Their own campaign of allegations is not in these circumstances a lawful exercise of free speech, much less an example of investigative journalism. It is unsubstantiated character assassination and it must stop.