{"id":200085,"date":"2026-08-09T16:42:26","date_gmt":"2026-08-10T00:42:26","guid":{"rendered":"https:\/\/lukeford.net\/blog\/?p=200085"},"modified":"2026-08-09T18:21:08","modified_gmt":"2026-08-10T02:21:08","slug":"heres-how-i-exposed-jason-arday-and-cambridge-david-starkey-talks-to-nathan-cofnas","status":"publish","type":"post","link":"https:\/\/lukeford.net\/blog\/?p=200085","title":{"rendered":"&#8220;Here&#8217;s how I exposed Jason Arday and Cambridge&#8221; | David Starkey Talks to Nathan Cofnas"},"content":{"rendered":"<p><iframe width=\"560\" height=\"315\" src=\"https:\/\/www.youtube.com\/embed\/C27UXFdz0sA?si=fZlHDa9bTyxyzMPL\" title=\"YouTube video player\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe><\/p>\n<p>0:00 David Starkey (b. 1945) opens by calling Nathan Cofnas (b. 1987) the David who brought down the monster of DEI at Cambridge.<\/p>\n<p>0:32 Both men name their motive as vengeance. Cambridge expelled Cofnas after two years, Starkey after fifty-six.<\/p>\n<p>1:20 Starkey credits X and social media, and says the BBC and the mainstream press would have buried the story.<\/p>\n<p>2:54 Cofnas gives his history: MPhil at Cambridge in 2015, DPhil at Oxford, return in 2022 as a Leverhulme Early Career Fellow on a three-year contract, plus a research associateship at Emmanuel College.<\/p>\n<p>3:31 He describes his field as philosophy of biology and divergence between ancestral populations. Starkey translates: race differences in intelligence.<\/p>\n<p>4:17 January 2024, he launches a Substack. The first essay argues the right has a stupidity problem.<\/p>\n<p>4:43 He cites the General Social Survey: respondents who attribute racial disparities to genetics score lowest on the abbreviated vocabulary and logic tests.<\/p>\n<p>6:20 Stereotype accuracy. He offers Jews and wealth, Jamaicans and sprinting.<\/p>\n<p>7:33 Nobody at Cambridge objected to that essay. No complaint about grading conservative students.<\/p>\n<p>8:29 The follow-up argues a coherent right requires hereditarianism.<\/p>\n<p>9:08 February 2024, &#8220;A Guide for the Hereditarian Revolution,&#8221; including the American case.<\/p>\n<p>10:15 Starkey presses on provenance. The Harvard number was not Cofnas&#8217;s own.<\/p>\n<p>11:26 The figure: Black enrollment around 12.5 percent, falling under grades and scores alone to under 0.7 percent.<\/p>\n<p>12:34 Cofnas says the White share would rise a couple of points and Asians would take most of the displaced places.<\/p>\n<p>13:08 He turned down an offer from an Asian university because Cambridge had adopted a strong free speech policy in 2020 after the Noah Carl firing.<\/p>\n<p>14:18 Starkey on the Regent House revolt: tolerant instead of respectful, the Chicago principles, the fall of Stephen Toope (b. 1958).<\/p>\n<p>16:20 Student protests, chants, the student paper, the Daily Mail chasing clicks.<\/p>\n<p>17:45 Bhaskar Vira named. Cofnas alleges administrators and the head of faculty met student protesters, then opened an investigation.<\/p>\n<p>19:04 The university hires a lawyer to decide whether hereditarianism is a protected philosophical belief under the Grainger test.<\/p>\n<p>21:46 Cofnas describes a three-hour interrogation and a hostile investigator.<\/p>\n<p>22:25 The report finds his belief fails Grainger and falls outside academic freedom for want of scholarly competence.<\/p>\n<p>23:02 He alleges the report treats 0.67 as larger than 0.7 and misses the count of Black Americans above IQ 135 by two orders of magnitude.<\/p>\n<p>24:14 Not sacked. Quarantined. Emmanuel later terminated the associateship.<\/p>\n<p>24:38 Support came privately. A letter to the Times, signed by a few, including Matthew Kramer (b. 1959), who called Cofnas an idiot while defending his speech.<\/p>\n<p>26:17 Starkey proposes revenge as the motive for the Arday audit. Cofnas denies it, then concedes the sequel appealed to him.<\/p>\n<p>27:13 A Cambridge whistleblower contacted him on June 13.<\/p>\n<p>27:39 He left Cambridge in September 2025, before the rumors became an open secret.<\/p>\n<p>28:10 Journalists had been deterred by legal threats.<\/p>\n<p>29:13 The dissertation: alleged copying from a then graduate student at Brunel.<\/p>\n<p>29:31 The biography as presented: autistic, nonverbal to eleven, unable to spell his name to eighteen, 600 miles in six days, thirty marathons in thirty-five days, over five million pounds raised, PhD to Cambridge chair by thirty-seven.<\/p>\n<p>31:03 Cofnas says recovery of normal speech after being nonverbal to eleven has no precedent he can find, and that two GCSEs still require writing your name.<\/p>\n<p>32:01 He feared a vexatious suit and confined himself to the published record.<\/p>\n<p>32:42 Copyleaks output, retained copyediting errors, a stray comma in the dissertation title.<\/p>\n<p>33:48 The heart of the piece is not the thesis. It is the ten items Arday lists on his own faculty page as principal publications, 2018 to 2023.<\/p>\n<p>34:43 His three tests for each: plagiarized, substantive, and actually his.<\/p>\n<p>35:27 Allegations of invented data, a &#8220;data analysis&#8221; section without statistics, and a convenience sample consisting of three friends.<\/p>\n<p>37:21 Starkey turns to the institution and says Cambridge swapped him out for Jason Arday. <\/p>\n<p>38:03 The appointment board was chaired by Sally Morgan (b. 1959), Blair&#8217;s former political director and Master of Fitzwilliam, the college that removed Starkey in five minutes.<\/p>\n<p>39:31 Labour threads: David Sainsbury (b. 1940) as former chancellor, Chris Smith (b. 1951) as current one.<\/p>\n<p>40:31 Cofnas names Simon Baron-Cohen (b. 1958) among Arday&#8217;s defenders.<\/p>\n<p>41:10 His thesis: wokeism rests on an empirical premise of identical innate group distributions, and everything follows from that premise plus Christian moral sentiment.<\/p>\n<p>44:24 Starkey pushes back. There are second-order causes, cultural and legal.<\/p>\n<p>45:16 His evidence for the shift: in 2009 he shared the Cambridge 800th anniversary American launch with Stephen Hawking and fronted the Open Day for free.<\/p>\n<p>46:26 His date is 2012. David Willetts (b. 1956) lifts the fee cap to \u00a39,000. The condition is access enforcement through OFFA under Les Ebdon (b. 1947). American DEI is imported wholesale onto that frame.<\/p>\n<p>48:30 The enabling statute comes from a Conservative government.<\/p>\n<p>49:36 Cofnas answers that law follows elite belief and that the ideology of the bureaucrat outranks the text.<\/p>\n<p>50:17 His long history: race denial crystallizes in the late nineteenth century, hardens after 1945, then the civil rights prediction fails and affirmative action ratchets upward for sixty years.<\/p>\n<p>53:08 Starkey adds Jonathan Haidt and the iPhone, and describes cancellation as heresy trial and execution in one motion.<\/p>\n<p>54:36 They agree DEI and scholarship cannot coexist.<\/p>\n<p>55:40 Starkey notes that Arday&#8217;s defenders argue the case damages the moral and practical case for DEI.<\/p>\n<p>56:28 The Korea thought experiment. Cofnas concedes a bonus for North Korean applicants would be rational. Starkey supplies East and West Germany.<\/p>\n<p>59:02 The claim that DEI requires assent to a proposition free speech would expose.<\/p>\n<p>1:00:15 The Chicago principles as remedy. Cofnas doubts they work, citing a Chicago department&#8217;s 2020 announcement restricting admissions to Black studies.<\/p>\n<p>1:02:20 Lionel Shriver (b. 1957) and Mania as the reductio.<\/p>\n<p>1:03:38 The most damaging line in the hour: thousands of scholars are as weak as Arday and simply do not plagiarize. Absent the copying and the biography, there would be no story.<\/p>\n<p>1:05:31 Optimism from information access. Cofnas learned of the literature at seventeen in a classroom. Zoomers found it in a search box.<\/p>\n<p>1:07:43 The Free Speech Union.<\/p>\n<p>Cofnas took the ten publications Arday nominated as his best work and asked three questions of each. That method needs no theory of population genetics, no view about the Grainger test, and no opinion about Cambridge. Anyone can run it against any professor, and the subject supplies the sample. That is a portable audit, and it is the part of this that will outlive the news cycle. It also worked. Arday resigned on August 5, 2026, hours after Cambridge opened a new investigation into his qualifications and honorary appointments, separate from the misconduct review already running. Jesus College opened its own. Cambridge had defended him in June, calling the campaign vile, and reversed within weeks.<\/p>\n<p>The second expos\u00e9 is an argument about why the first was necessary, and it runs on assertion. Cofnas says at 36 and 43 that Arday is mentally disabled, could not do addition, is mathematically illiterate. None of that is in the documents. He is entitled to think it. Putting it in the same hour as the plagiarism analysis hands the defense its best material, and the Good Law Project took it. Their petition frames the plagiarism claims as fake and racially motivated, and Arday told the Times he had made mistakes and should be held accountable while denying he is a liar and a fraud. A finding that survives independent replication is one thing. A characterization of a man&#8217;s intelligence is not replicable, and it converts an audit into a fight about the auditor.<\/p>\n<p>Starkey&#8217;s date is 2012 and his causal chain runs from Willetts through the fee cap to OFFA to Ebdon to imported DEI. The chain is off. OFFA was created by the Higher Education Act 2004, under Labour, alongside the first \u00a33,000 cap. Willetts raised the cap to \u00a39,000 and Ebdon took the directorship in 2012, so what changed that year was the price and therefore the leverage. The weaker version of Starkey&#8217;s claim survives and is better: an access regulator with little to withhold became an access regulator with a great deal to withhold, and the compliance apparatus grew to match. That is a testable proposition about access agreements before and after 2012, and nobody in the hour tests it.<\/p>\n<p>Grainger plc v Nicholson (2010) sets five criteria for a protected philosophical belief, one of which excludes an opinion based on the present state of information available. Forstater v Center for Global Development (2021) then narrowed the exclusion so that only beliefs approaching Nazism or totalitarianism fall outside protection. A university that promises to protect all lawful speech and then asks whether the speech is separately protected under employment discrimination law has changed the question. That substitution is the most institutionally significant thing in the interview and gets four minutes.<\/p>\n<p>The line at 1:03:38 undoes the frame Starkey spent an hour building. If thousands of scholars produce work as thin as Arday&#8217;s and are safe because they do not copy, then plagiarism is incidental and the real charge is against a field&#8217;s standards of evidence. That charge is larger, harder, and needs different work: sampling published articles in the sociology of education, coding for method, showing what a convenience sample of three friends yields relative to what the field accepts. Cofnas gestures at it and moves on. Nobody has done it. Someone should.<\/p>\n<p>What the hour never does is ask what would falsify anything. Starkey supplies the vengeance frame in the first minute and Cofnas accepts it with a grin, which means every claim that follows arrives pre-labeled with a motive. Liverpool John Moores investigated the thesis and attributed the problems to honest and reasonable error. That finding is not mentioned. Neither is any account of what Arday&#8217;s supervisors and examiners did, which is the question Cambridge cannot answer by investigating one man. Simon &#038; Schuster still plans to publish his memoir this month.<\/p>\n<p><iframe width=\"560\" height=\"315\" src=\"https:\/\/www.youtube.com\/embed\/JKQePI1CfNg?si=JWe45Kqdi5gfSA-J\" title=\"YouTube video player\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe><\/p>\n<p>2:16 to 2:52. Cofnas says he argued in January 2024 that conservatives, his own group, have lower average IQ for genetic reasons, and that Cambridge said nothing. A month later he wrote the hereditarian piece and the investigation began. He is building an argument about selective enforcement, and the setup is the strongest rhetorical move in the hour. He applies his framework to the tribe he belongs to.<\/p>\n<p>7:59 to 9:56. The method, which is the part worth stealing. He ran the dissertation through Copyleaks. Then he took Arday&#8217;s own faculty page, the section listing principal publications, and audited each of the ten against three tests: is it plagiarized, does it contain scholarly content, did he write it. Using the subject&#8217;s own self-selected best work as the sample removes the cherry-picking objection.<\/p>\n<p>10:20 to 11:39. Arday called the police on Dave Harris, the academic who raised questions in 2023, and on Jack Grove of Times Higher Education. Carter-Ruck sent a letter to the journalist. This is the part of the story with the most legs and the least coverage.<\/p>\n<p>14:33 to 15:55. The framing war stated openly. Grove&#8217;s angle is research misconduct that could have happened to anyone. Cofnas&#8217;s angle is DEI. Both men are describing the same documents. Cofnas calls it the most extreme case of plagiarism in academia, which is the kind of superlative that invites a counterexample.<\/p>\n<p>17:34 to 19:33. The reversal. Cofnas says he would remove responsibility from Arday and offers the bar-braggart analogy: if I appoint a man boasting in a pub to a Cambridge chair, the fault is mine. Marshall pushes. Within a minute Cofnas has Arday morally culpable, aware he was stealing, and careful to plagiarize only people without institutional power, including a graduate student at Brunel. Then the line that undoes his own analogy: &#8220;he&#8217;s not as mentally disabled as he claims.&#8221; Strategic plagiarism requires the calculation the bar-drunk framing denies.<\/p>\n<p>20:32 to 21:05. Simon Baron-Cohen (b. 1958) signed the Good Law Project letter and defended Arday on Twitter on autism grounds. Cofnas calls this the most surprising signature and declines to speculate about why. He speculates freely everywhere else.<\/p>\n<p>21:13 to 24:28. The best transferable idea. Signing the letter costs almost nothing, since the worst case is mild embarrassment on a list of prominent names. Opposing carries real career risk if the consensus holds. So the expected value runs one way. That accounts for the signatures without requiring anyone to believe anything.<\/p>\n<p>35:32 to 40:09. The buried lede. Marshall spends twenty minutes trying to get the race realist to endorse colorblindness and cannot. Cofnas advocates open racial quotas. Set aside a stated percentage, choose the best candidates within it, and stop pretending. His argument is that colorblindness plus real gaps produces representation so low that no multiracial society will accept it, so the choice is between honest quotas and the current system of personality scores and denial.<\/p>\n<p>40:25. The crime and prisons analogy for why his stated preference does not govern his policy.<\/p>\n<p>51:38 and 55:54. The load-bearing claim. If the root is genetic, money cannot close the gap, and the Flynn effect has &#8220;basically maxed out,&#8221; and we are &#8220;very close to&#8221; an environment where people reach their genetic potential. He asserts this. Marshall lets it go.<\/p>\n<p>1:04:17 to 1:07:15. Antisemitism as an application of the same logic he attributes to wokeness: a group does better, sameness is assumed, so the success must be theft. He is making the hereditarian case as a defense of Jews and of overseas Chinese and of Indians in Africa. Whatever else it is, it is consistent.<\/p>\n<p>1:13:02 to 1:15:42. Marshall catches the slide. Cofnas said conservatives have lower IQ for genetic reasons. Under pressure it becomes people who currently identify as conservative, then &#8220;there isn&#8217;t like the definitive study,&#8221; then the General Social Survey, then indirect evidence that conservatives cannot build institutions. The GSS instrument is a ten-word vocabulary test. The claim that started his troubles rests on it.<\/p>\n<p>1:16:07 to 1:18:26. The Emmanuel College position was unpaid and terminated. The three-year university contract expired. He distinguishes the two carefully, which the coverage has not. The Ghent yogurt attack, the identified assailant, and the prosecutor declining to act.<\/p>\n<p>Now the thoughts.<\/p>\n<p>The story nobody will write is that the man the Guardian files under far-right race realism spent half this interview advocating racial quotas. Not colorblindness. Quotas, stated openly, applied to Cambridge chairs and Harvard admissions. He reaches affirmative action from the hereditarian premise, and Marshall, who wants him to say something else, keeps handing him openings he declines.<\/p>\n<p>In &#8220;Pursuing Diversity&#8221; Amy Wax argues that shifting the rationale from remediation to diversity removes the endpoint, so preference becomes permanent, and she offers that as a reductio. Cofnas reaches the same destination and calls it policy. If the gaps are genetic, no remedy closes them, so any society that wants representation needs permanent preference. Affirmative action forever, from opposite premises, by two people who would each find the other&#8217;s route intolerable. Neither cites the other. <\/p>\n<p>The empirical test has now run and Marshall missed it. Cofnas rests everything on Harvard at 0.7 percent under colorblind admissions. SFFA ended race-conscious admissions in 2023. Harvard&#8217;s class of 2029 came in at 11.5 percent Black, down from 14 percent. Either Harvard is using proxies to reproduce the old outcome, which Cofnas would say and which the Court may eventually test, or the simulation was never a good prediction of what colorblind admissions produce. The gap between 0.7 and 11.5 is the whole argument, and nobody in the room raised it.<\/p>\n<p>The Flynn passage is where his own standard fails. He demands that everyone else show the magnitude of an effect and not merely its existence. Then he asserts that environmental intervention has approached its limit, which is the claim doing all the work in his argument, and gives no evidence for it. Apply the test he applies to others and the hereditarian case has the same shape as the case he attacks: a real phenomenon, an unmeasured share of the variance, and a policy conclusion arriving early.<\/p>\n<p>The Fryer passage. Marshall says trumped-up charges, Cofnas agrees, and the segment proceeds as though Roland Fryer (b. 1977) was punished for the policing paper. Harvard&#8217;s stated basis was a Title IX finding on unwelcome conduct. You can argue the sanction was disproportionate or that the investigation would not have happened to a differently positioned man. Repeating &#8220;obviously motivated by political considerations&#8221; as established fact costs you credibility. <\/p>\n<p>Baron-Cohen is the unwritten story. The world&#8217;s most cited autism researcher publicly defends a man on autism grounds, and the autism-adjacent claims in that man&#8217;s biography, nonverbal until eleven and unable to read until eighteen, are among the ones most in doubt. Cofnas notices and declines. I have asked Professor Baron-Cohen what evidence he relied on. <\/p>\n<p>Marshall (b. 1987) does both things in one hour. His short neutral questions produce the best material: why does that matter, can you explain to me why, what would be an example of a positive thing to have come out of this. Then from about 1:07 he starts announcing his own position, &#8220;you still haven&#8217;t won me over,&#8221; &#8220;I&#8217;m not persuaded,&#8221; and the last fifteen minutes turn into a debate where Cofnas defends instead of describes. Compare what he gets before and after that switch. It is the clearest demonstration I have seen recently of why questions should not carry a moral judgment or thesis. <\/p>\n<p>The interview posted August 8 and the events moved under it. Arday resigned from the university and from Jesus College on August 5, hours after Cambridge announced an investigation into his qualifications and honorary appointments, with separate misconduct complaints proceeding under the research policy. He denies wrongdoing and says the resignation is not an admission. He told the Times he made academic mistakes and that the campaign against him was racially motivated. Jesus College opened its own inquiry. Simon &#038; Schuster&#8217;s 37 Ink still has his memoir scheduled. Cofnas speaks in the transcript as though the quitting is settled and the investigations are not, which is right, and the story now has three separate proceedings running rather than one.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>0:00 David Starkey (b. 1945) opens by calling Nathan Cofnas (b. 1987) the David who brought down the monster of DEI at Cambridge. 0:32 Both men name their motive as vengeance. Cambridge expelled Cofnas after two years, Starkey after fifty-six. &hellip; <a href=\"https:\/\/lukeford.net\/blog\/?p=200085\">Continue reading <span class=\"meta-nav\">&rarr;<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_monsterinsights_skip_tracking":false,"footnotes":""},"categories":[16281,43259,42838],"tags":[],"class_list":["post-200085","post","type-post","status-publish","format-standard","hentry","category-college","category-jason-arday","category-nathan-cofnas"],"_links":{"self":[{"href":"https:\/\/lukeford.net\/blog\/index.php?rest_route=\/wp\/v2\/posts\/200085","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lukeford.net\/blog\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/lukeford.net\/blog\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/lukeford.net\/blog\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/lukeford.net\/blog\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=200085"}],"version-history":[{"count":8,"href":"https:\/\/lukeford.net\/blog\/index.php?rest_route=\/wp\/v2\/posts\/200085\/revisions"}],"predecessor-version":[{"id":200093,"href":"https:\/\/lukeford.net\/blog\/index.php?rest_route=\/wp\/v2\/posts\/200085\/revisions\/200093"}],"wp:attachment":[{"href":"https:\/\/lukeford.net\/blog\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=200085"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/lukeford.net\/blog\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=200085"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/lukeford.net\/blog\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=200085"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}